McKinsey>60% from agentic
Agentic AI carries most of the value
More than 60% of the increased value from AI in marketing and sales is expected to come from agentic deployments rather than assistive ones.
Program 17 · Domain specialisation
SCIKIQ Certified Data and AI EngineerMarketing, Sales & Service
Commercial functions are where agentic AI has the largest measured value pool and the worst signal-to-noise. Everything demos well; very little survives contact with a real CRM. The difference is almost always data quality, consent and whether the agent can actually act in the system of record.
- 80 hrstaught
- 40hands-on AI labs
- 8assessment tests
- +20 hrs2 E2E capstones
>60%Of AI value in marketing and sales expected to come from agentic AIMcKinsey
71% vs 25%High-growth firms raising AI investment by double digits versus the restMcKinsey
$2.6-4.4tnAnnual value gen AI could unlock, with marketing and sales a leading poolMcKinsey
4,000Buyers and sellers surveyed in the 2026 B2B Pulse across 13 countriesMcKinsey
01
Part 1 · The business
The value chain, the people, the trends
The commercial chain runs from demand creation to renewal. Part 1 maps it as the impact journeys the strategy houses now use, because that is the unit of rewiring that actually pays.
The commercial value chain
Six stages from demand to advocacy. Select one.
01
Create demand
Brand, content, campaigns, SEO and demand generation.
Decisions made here
- Who do we target?
- With what message?
- On which channel?
Where it breaks today
- Content production is the bottleneck
- Personalisation that is really segmentation
- Attribution nobody believes
What is measured
- Pipeline created
- Cost per lead
- Engagement rate
Where AI landsContent generation at scale, journey orchestration and grounded audience selection.
Pipeline createdCost per leadEngagement rate
02
Capture & qualify
Inbound handling, lead scoring, routing and qualification.
Decisions made here
- Is this lead real?
- Who should own it?
- Is it ready for sales?
Where it breaks today
- Slow response to inbound interest
- Scoring models nobody trusts
- Sales and marketing disagreeing about quality
What is measured
- Speed to lead
- Conversion to opportunity
- Lead acceptance rate
Where AI landsInstant inbound engagement, enrichment-based qualification and explainable routing.
Speed to leadConversion to opportunityLead acceptance rate
03
Sell & advance
Discovery, solutioning, proposals, negotiation and close.
Decisions made here
- What does this buyer need?
- What do we propose?
- What is blocking the deal?
Where it breaks today
- Sellers spending most of their week not selling
- Proposal assembly done manually
- Deal risk visible only in hindsight
What is measured
- Win rate
- Sales cycle length
- Selling time
Where AI landsAccount research, proposal drafting, meeting preparation and deal-risk surfacing.
Win rateSales cycle lengthSelling time
04
Price & quote
Configuration, pricing, approvals and contracting.
Decisions made here
- What price and terms?
- Does this need approval?
- Is this configuration valid?
Where it breaks today
- Quote cycles measured in days
- Discounting without evidence
- Approval chains that stall deals
What is measured
- Quote turnaround
- Discount leakage
- Approval cycle time
Where AI landsGuided configuration, evidence-based pricing guidance and drafted contract terms.
Quote turnaroundDiscount leakageApproval cycle time
05
Serve & support
Customer service, technical support, self-service and case management.
Decisions made here
- What is the customer asking?
- Can this be resolved now?
- Does this need escalation?
Where it breaks today
- Repeat contacts from partial answers
- Knowledge scattered across systems
- Agents reading rather than resolving
What is measured
- First-contact resolution
- Handle time
- CSAT
- Deflection
Where AI landsGrounded resolution, action-taking under approval and knowledge gap detection.
First-contact resolutionHandle timeCSATDeflection
06
Retain & grow
Onboarding, adoption, renewal, expansion and advocacy.
Decisions made here
- Is this customer healthy?
- What is the expansion play?
- Are we at renewal risk?
Where it breaks today
- Churn signals spotted after the decision
- Expansion identified inconsistently
- Success teams working from stale data
What is measured
- Net revenue retention
- Churn rate
- Expansion rate
Where AI landsHealth scoring with reasons, renewal risk detection and expansion play recommendation.
Net revenue retentionChurn rateExpansion rate
Who you are building for
Commercial users abandon tools faster than any other group. Adoption is the design problem. Select one to light the stages they own.
What the strategy houses are seeing
The largest measured value pool in enterprise AI — and the noisiest market.
McKinseyJourneys, not tools
Rewire journeys, not tools
Growth leaders rewire commercial “impact journeys” end to end — pricing, sales cycle, engagement, cost to serve — rather than adding point tools.
McKinsey71% vs 25%
Investment separates the growers
In the 2026 B2B Pulse, 71% of high-growth companies increased AI investment by double digits versus 25% of the rest.
McKinseyNew buying surface
Agentic commerce changes discovery
When agents perform discovery on behalf of buyers, machine-readable content and structured product data become acquisition infrastructure.
BainOne record
Unified platforms beat point solutions
Fragmented agent tooling produces fragmented data. The pattern that scales runs agents against one governed customer record.
HBRWrong > generic in harm
Trust degrades fast with bad personalisation
Personalisation built on wrong or stale data damages the relationship more than no personalisation at all.
Where these numbers come from — 4 sources
Consulted when this program was written, in July 2026. Each entry names the claim on this page that it backs, so the pairing stays checkable if the copy is later edited.
- McKinseyAgents for growth: turning AI promise into impact>60% of marketing and sales AI value from agentic deployments
- McKinseyThe future of B2B sales: how growth champions rewire their playbooks with AI71% vs 25% AI investment increase, 2026 B2B Pulse
- McKinseyReinventing marketing workflows with agentic AIRewiring impact journeys rather than adding tools
- IndustryAI sales agents: the 2026 category guideUnified platforms and governed customer records
02
Part 2 · Data and meaning
Domains, ontology, taxonomy and governed terms
The commercial stack is the most fragmented in the enterprise. The architecture problem is producing one trustworthy customer record that an agent can both read and safely write to.
The data landscape
Six families that rarely agree about who the customer is.
CRM
Accounts, contacts, opportunities and activity — the system of record, and often the weakest data.
Salesforce, Dynamics, HubSpotMarketing automation
Campaigns, journeys, engagement and consent state.
Marketo, Braze, HubSpotProduct & usage
Product telemetry, adoption and entitlement — the best churn predictor you have.
Product analytics, entitlementsService & support
Cases, tickets, knowledge base and conversation history.
Zendesk, ServiceNow, SalesforceEnrichment & intent
Firmographic, technographic and third-party intent signals.
Enrichment vendors, intent dataCommercial content
Approved messaging, collateral, pricing rules and contract templates.
CMS, CPQ, CLMThe semantic layer
The commercial stack is the most fragmented in the enterprise and the one where a wrong identity does the most damage. Resolving party, account and interaction into one governed graph is what makes personalisation safe rather than embarrassing.
Data domains
Six subject areas, each with the entities it holds, its critical data elements and the function accountable for it.
Party & identity
People and organisations, resolved across every touchpoint.
PersonAccountContactHouseholdIdentity link
- Critical data elements
- Party identifier; Match confidence; Golden record flag
- Owned by
- Revenue operations
Consent & preference
What you are permitted to do with each party, by channel and purpose.
ConsentSuppressionPreferenceLawful basis
- Critical data elements
- Consent state; Channel; Captured timestamp
- Owned by
- Privacy & RevOps
Interaction
Every touch: campaign, conversation, visit, case.
CampaignActivityConversationCaseWeb session
- Critical data elements
- Interaction timestamp; Channel; Outcome
- Owned by
- Marketing & service
Pipeline & revenue
Opportunities, quotes, contracts and the revenue they become.
LeadOpportunityQuoteContractSubscription
- Critical data elements
- Stage; Amount; Close date; Renewal date
- Owned by
- Sales operations
Product & entitlement
What is sold, how it is priced and what the customer is entitled to.
ProductPrice bookEntitlementUsage
- Critical data elements
- Product code; Contracted quantity; Usage metric
- Owned by
- Product & CPQ
Content & claims
Approved messaging and the assets built from it.
Approved claimAssetKnowledge article
- Critical data elements
- Claim identifier; Market scope; Expiry date
- Owned by
- Marketing & compliance
Ontology & knowledge graph
The entities and the typed relationships between them — what the agent traverses instead of guessing joins. Select any entity.
Core entityEventReference / classification
Core entity
Person
An individual, resolved across systems.
- Person works at Account
- Person grants Consent
- Campaign targets Person
- Case raised by Person
Core entity
Account
An organisation you sell to or serve.
- Person works at Account
- Opportunity held by Account
- Contract entitles Account
Reference / classification
Consent
Permission to contact, by channel and purpose.
- Person grants Consent
Event
Campaign
A coordinated outbound programme.
- Campaign targets Person
- Campaign generates Activity
- Campaign uses Approved claim
Event
Activity
A single interaction with a person.
- Activity creates Lead
- Campaign generates Activity
Event
Lead
An expression of interest not yet qualified.
- Lead converts to Opportunity
- Activity creates Lead
Event
Opportunity
A qualified potential sale.
- Opportunity held by Account
- Opportunity priced as Quote
- Lead converts to Opportunity
Event
Quote
A priced, configured proposal.
- Quote contains Product
- Quote becomes Contract
- Opportunity priced as Quote
Core entity
Contract
The executed agreement and its entitlements.
- Contract entitles Account
- Quote becomes Contract
Reference / classification
Product
The offering being sold.
- Quote contains Product
- Case concerns Product
Event
Case
A service request raised by a customer.
- Case raised by Person
- Case concerns Product
Reference / classification
Approved claim
A statement compliance has cleared for use.
- Campaign uses Approved claim
Taxonomies
The classification hierarchies that make records comparable across systems.
Account hierarchy
- Global parent
- Region
- Legal entity
- Site
GlobalCo → EMEA → GlobalCo GmbH → Munich office
Interaction & channel
- Type
- Channel
- Sub-channel
Outbound → Email → Nurture sequence
Case classification
- Category
- Type
- Reason
Technical → Integration → Authentication failure
Governed business terms
The definitions an agent must use rather than invent. Most wrong answers in this industry are a term used loosely.
- Lead vs contact
- A lead is unqualified interest; a contact is a known person on an account. Merging them corrupts conversion metrics.
- Account hierarchy
- The parent-child structure of an organisation. Revenue rolls up it; entitlements usually do not.
- Consent
- Permission to contact, specific to channel and purpose, and valid only until withdrawn.
- Suppression
- A rule preventing contact regardless of consent — it always wins.
- Qualified opportunity
- A potential sale meeting agreed criteria. The criteria must be explicit or the pipeline is fiction.
- Net revenue retention
- Retained plus expanded revenue from existing customers, excluding new logos.
- First-contact resolution
- Cases resolved in one interaction with no follow-up contact within a defined window.
- Approved claim
- A statement compliance has cleared for use in a specific market. Anything else is unapproved.
03
Part 2 · The architecture
How it is actually built
The layers a deployment needs, what each one holds, and what the engineer owns there.
The reference architecture
Seven layers built around one governed customer record.
01/07
Sources & systems
CRM, marketing automation, product, service and enrichment sources.
Takes in
Nothing. This is the upstream edge.
Hands on
Identity-resolved profiles, interaction history and the consent state attached to each.
SystemWhat it doesKey business entities
CRM
Accounts, contacts, pipeline and cases. The system sales and service actually work in, and the one whose data quality reflects that.
AccountContactLeadOpportunityCase
Marketing automation and CDP
Campaigns, journeys, segments and the identity graph that stitches a person across devices and channels.
PersonCampaignActivityIdentity link
Consent and preference management
Lawful basis, consent state, suppression lists and preferences by channel and purpose. Governs whether an action is permitted at all.
ConsentSuppressionPreferenceLawful basis
Service desk and contact centre
Cases, conversations, transcripts and knowledge articles across voice and digital — the largest unstructured corpus most commercial teams own.
CaseConversationKnowledge article
Commerce and billing
Quotes, contracts, subscriptions, entitlement and usage. What the customer is actually owed and actually using.
QuoteContractSubscriptionEntitlementUsage
Approved content and claims
Brand assets, approved claims and regulated messaging. Determines what an agent is permitted to say, as distinct from what it can retrieve.
Approved claimAssetBrand rule
What the engineer owns here
Establishing which system is authoritative for each attribute, in writing.
Where this layer fails
Resolving identity before checking consent. A beautifully stitched profile you are not permitted to act on is a liability, not an asset — consent travels with the identity or the design is wrong.
02/07
Ingestion & identity resolution
Streaming and batch ingestion with identity resolution across every touchpoint.
Takes in
Read-mostly access paths, change feeds and document streams from the systems of record.
Hands on
Replayable, resolved, quality-checked records with their lineage back to the source row.
SystemWhat it doesKey business entities
Change data capture
Reads the source’s own change log rather than polling it, so the platform sees every state a record passed through instead of only where it ended up.
Change eventOperation typeCommit timestamp
Landing and raw zone
Every extract kept as it arrived, immutable and timestamped. The thing you replay from when a downstream definition turns out to have been wrong.
Raw extractBatchWatermark
Document processing
Layout-aware extraction over the unstructured half of the estate: sectioning, tables, signatures, and the page reference every later citation depends on.
DocumentSectionPage reference
Entity resolution
Decides that two records are the same real-world thing, with a survivorship rule and a confidence, so downstream joins are a decision rather than an assumption.
Match candidateSurvivorship ruleCluster
Data quality and contracts
Schema, freshness and volume expectations asserted at the boundary, so a bad load fails loudly here rather than quietly two layers later.
ContractExpectationViolation
What the engineer owns here
Identity resolution — the layer everything else depends on and nobody funds.
Where this layer fails
Snapshot loads instead of change capture. It looks identical in a demo and it silently loses every intermediate state, which is exactly what an audit later asks you for.
03/07
Governance, consent & semantics
The governed customer record, consent state, suppression rules and governed metrics.
Takes in
Replayable, resolved records with lineage from ingestion.
Hands on
Typed, classified, defined data an agent can be grounded on without inferring meaning.
SystemWhat it doesKey business entities
Business glossary
One agreed definition per term, owned by a named person, so the number an agent quotes means what the business means by it.
TermDefinitionOwnerSteward
Ontology and knowledge graph
The typed entities and relationships the domain actually has, so an agent can traverse "which customers are exposed to this" rather than guess from adjacent text.
EntityRelationshipNodeEdge
Semantic and metrics layer
Metrics defined once, in one place, with their filters and grain. Removes the class of error where the agent computed something plausible and wrong.
MetricDimensionGrainFilter
Classification and policy
Sensitivity labels and the purpose each classification permits, applied at the field level and inherited by everything downstream.
ClassificationPurposePolicy
Lineage and catalogue
Where a value came from and what it touched on the way. The layer that makes an answer defensible rather than merely correct.
Lineage edgeDatasetColumn
What the engineer owns here
Consent as an enforced gate, not a documented policy.
Where this layer fails
Letting the model infer meaning from column names. It will, it will be plausible, and nobody will notice until the number reaches a regulator or a board pack.
04/07
Retrieval & grounding
Retrieval over approved messaging, product content, case history and account intelligence.
Takes in
Typed, classified, defined data from the governance layer.
Hands on
Ranked, permission-trimmed, citable evidence scoped to the caller and the moment.
SystemWhat it doesKey business entities
Hybrid index
Vector and lexical retrieval together, because exact identifiers, codes and clause numbers are the thing semantic search is worst at.
ChunkEmbeddingTermScore
Chunking and metadata
Splits on the document’s own structure and carries effective dates, version and source into every chunk, so a retrieved passage knows when it was true.
ChunkEffective dateVersionSource reference
Security trimming
Applies the caller’s permissions inside the query rather than filtering results afterwards, so the model never sees what the user may not.
PrincipalEntitlementFilter
Graph retrieval
Traverses the knowledge graph for questions that are joins rather than similarity — exposure, genealogy, ownership, causation.
PathHopTraversal
Reranking and citation
Reorders candidates on relevance and returns the passage identifier behind every sentence, so the answer can be checked rather than trusted.
CandidateRankCitation
Consent-aware retrievalthis domain adds
Applies suppression and lawful basis inside the query, so a profile that may not be acted on is not retrievable for that purpose in the first place.
Consent stateSuppressionPurpose
What the engineer owns here
Ensuring generated messaging comes from approved claims, not model imagination.
Where this layer fails
Retrieving on identity and checking consent at send time. By then the profile is in the context window and the model has already reasoned over it.
05/07
Agent & orchestration
Content agents, inbound engagement, meeting preparation, service resolution and health scoring.
Takes in
Ranked, permission-trimmed, citable evidence from retrieval.
Hands on
Actions taken or proposed, each with the evidence, the identity and the trace behind it.
SystemWhat it doesKey business entities
Agent runtime
Plans, calls tools and holds the loop. Where the step budget, the timeout and the stopping condition are enforced rather than hoped for.
TaskPlanStepTrace
Tool and MCP servers
Typed, permissioned tools with declared schemas. An agent’s real capability surface is this list, which is why the list is a security artefact.
ToolSchemaScopeInvocation
State and memory
Durable task state with checkpoints, so a run that dies mid-way resumes instead of restarting and re-doing side effects.
CheckpointSessionMemory record
Human-in-the-loop gates
Approval steps on the actions that need one, carrying enough context for the approver to actually decide rather than rubber-stamp.
ApprovalThresholdDecision
Workflow integration
Writes back into the systems people already work in, under a service identity with its own audit trail.
Work itemWrite-backActor
What the engineer owns here
Write discipline — agents that update CRM without degrading it.
Where this layer fails
An agent that acts under a service account rather than on behalf of the user. It will eventually do something the requesting user had no right to do, and the log will not show it.
06/07
Guardrails, brand & privacy
Brand and claim compliance, consent enforcement, privacy controls and escalation rules.
Takes in
Proposed actions and drafted responses from the agent layer.
Hands on
Permitted, grounded, logged output — or a refusal with a stated reason.
SystemWhat it doesKey business entities
Policy enforcement
The rules that decide whether an action is permitted at all, evaluated before the action and independently of the model that proposed it.
PolicyDecisionObligation
Input and output screening
Injection detection on the way in, and on the way out the checks for leakage, unsupported claims and content the domain forbids.
PromptResponseViolation
Groundedness checking
Verifies each assertion resolves to retrieved evidence, and fails the response rather than shipping the sentence that does not.
AssertionSupportVerdict
Model risk management
Model inventory, intended use, validation evidence and the sign-off that lets a model be used for a purpose. Not optional in a regulated estate.
ModelIntended useValidationApproval
Audit log
Immutable record of what was asked, retrieved, decided and done — the artefact a reviewer reads when they do not take your word for it.
Audit entryActorEvidence reference
Approved-claim enforcementthis domain adds
Restricts what an agent may assert to the claim library approved for that product and market, which is narrower than what it can retrieve.
Approved claimMarketProduct
What the engineer owns here
Blocking outbound contact that consent does not permit, at the tool layer.
Where this layer fails
Guardrails that only run on the prompt. Most real failures are on the output side: the unsupported sentence, the leaked field, the claim nobody approved.
07/07
Evaluation & measurement
Conversion and resolution measurement, holdout design, data-quality monitoring and cost per interaction.
Takes in
Permitted, grounded, logged output from the guardrail layer.
Hands on
Measured quality, cost and latency — and the evidence to change any of the three.
SystemWhat it doesKey business entities
Evaluation harness
Held-out sets and graded runs on every change, so a prompt edit is a measured change rather than a hopeful one.
Eval setCaseScoreBaseline
Regression gate
Blocks a release when quality drops, in CI, on the same evidence for everyone. The difference between a system and a demo.
ThresholdRunGate decision
Tracing
End-to-end spans across retrieval, model and tool calls, so a bad answer can be opened and read rather than argued about.
TraceSpanLatencyToken count
Cost and capacity
Cost per task and per tenant, against throughput and quota. The number that decides whether the pilot can become the rollout.
Cost per taskQuotaThroughput
Drift and feedback
Watches quality against production traffic rather than the test set, and routes real corrections back into the eval suite.
Drift signalFeedbackIncident
What the engineer owns here
Measuring incremental conversion with a holdout, not attributed conversion.
Where this layer fails
Evaluating once, before launch. Quality moves with the data, the model and the traffic, and a system with no live measurement has no idea which of the three moved.
Typical tech stack
Models
- Frontier models for reasoning and generation
- Smaller models for classification and routing
- Voice models for service
Grounding
- Approved content and claims index
- Case and conversation history
- Account intelligence
Agents
- Inbound engagement agents
- MCP over CRM and service
- Journey orchestration
Data
- Identity resolution
- Governed customer record
- Consent registry
Assurance
- Holdout-based measurement
- CRM data quality monitoring
- Claim compliance checks
Runtime
- Consent enforcement at tool layer
- Write validation
- Full interaction retention
Constraints that shape the design
Consent is a hard gate
GDPR, CCPA and channel-specific rules govern outbound contact. Enforce consent in the tool, never in the prompt.
Do not corrupt the CRM
An agent writing low-quality data into the system of record destroys value faster than it creates it. Validate every write.
Identity resolution first
Personalisation on a wrong identity is worse than generic messaging. Resolution quality caps everything downstream.
Measure with holdouts
Attribution in commercial functions is contested. A holdout is the only number people accept.
04
Part 3 · Getting it done
Where the business and the technology meet
These twelve use cases span demand through renewal. The ones that work share a trait: the agent acts in the system of record under approval, on a governed customer record.
Use cases, mapped to the chain and the architecture
Filter by stage or by earned autonomy. Selecting a use case jumps the value chain and the architecture to the stage and layer it depends on.
Value chain stage
Autonomy earned
12 use cases
No use cases match that combination.
The delivery playbook
How a commercial deployment actually runs.
01
Week 1 — audit the customer record
Check identity resolution and consent state before promising personalisation. This determines what is possible.
02
Week 2 — pick one journey
One journey, one segment, one measurable outcome. Commercial deployments fail by trying to improve everything at once.
03
Weeks 3-5 — ground in approved content
Build the approved-claims index and the compliance gate before generating anything customer-facing.
04
Weeks 6-8 — design the holdout
Set up the holdout before launch. Retrofitted measurement will be argued with and dismissed.
05
Weeks 9-12 — protect the CRM and hand over
Instrument write quality, prove the record did not degrade, and hand over with monitoring.
The curriculum
8 modules, 80 taught hours, 40 hands-on labs and 8 assessments — every lab provisioned and graded by the SCIKIQ Agentic AI Playground. Open a module to see its labs.
01The SCDAI delivery model in commercial functionsLanding in a function that adopts fast, abandons faster and argues about attribution.10 hrs · 5 labs
Delivery mandateImpact journey selectionValue sizingStakeholder mapAdoption dynamicsThin-slice scoping
01
Journey selectionScore four impact journeys on value, data readiness and adoption risk.
02
Value sizingSize a conversion or resolution improvement in revenue terms.
03
Record auditAssess identity resolution and consent readiness for a use case.
04
Stakeholder mapMap sales, marketing, service, RevOps and privacy.
05
Thin sliceScope a one-journey slice with a holdout designed in.
Assessment 01 — Operating model & scoping
02Commercial data foundationsCRM, marketing, product and service data — and the identity problem underneath.10 hrs · 5 labs
CRM data modelsIdentity resolutionConsent stateProduct telemetryEnrichment dataData quality
01
Identity resolutionResolve customers across CRM, product and service with match scoring.
02
Consent registryBuild a consent store and enforce it at query time.
03
Product telemetryJoin usage data to accounts for health scoring.
04
EnrichmentIntegrate third-party enrichment and handle conflicts.
05
Quality monitoringDetect and alert on CRM data degradation.
Assessment 02 — Commercial data landscape
03Context engineering for customer-facing textEverything generated here may be read by a customer. Brand and claims are hard constraints.10 hrs · 5 labs
Brand voiceClaim complianceStructured outputsPersonalisation groundingRefusal and escalationContext budget
01
Brand groundingGenerate on-brand variants constrained by approved messaging.
02
Claim checkingBlock any generated claim not present in the approved library.
03
Personalisation groundingEnsure personalised facts come from the governed record.
04
EscalationMake the agent hand off to a human at the right moment.
05
Context budgetReduce cost on a high-volume conversational agent.
Assessment 03 — Context engineering
04Grounding on approved content and historyApproved messaging, case history and account intelligence as retrievable context.10 hrs · 5 labs
Approved content indexingCase history retrievalPrior-win retrievalFreshnessCitationsRetrieval evals
01
Content indexIndex approved messaging with version and market scoping.
02
Case retrievalRetrieve prior resolutions for a new service case.
03
Prior winsRetrieve comparable closed deals for proposal drafting.
04
FreshnessEnsure stale content is never used in a customer-facing answer.
05
Retrieval evalsMeasure resolution accuracy on labelled service cases.
Assessment 04 — Retrieval & grounding
05Agent design for commercial workflowsConversational agents, action-taking under approval, and not corrupting the CRM.10 hrs · 5 labs
Workflow vs agentConversational designHandoffCRM write disciplineMulti-agent commercial flowsContainment
01
Inbound agentBuild a qualifying conversational agent with graceful handoff.
02
Service resolutionResolve a case and execute the fix through an approved tool.
03
CRM write validationValidate every agent write against a schema and quality rules.
04
Multi-agent flowCoordinate research, drafting and update agents.
05
ContainmentScope agent action to the specific customer and permission set.
Assessment 05 — Agent design & autonomy
06Integrating with the commercial stackCRM, marketing automation, CPQ and service platforms — and their API limits.10 hrs · 5 labs
MCP over CRMMarketing platform integrationCPQ integrationAPI limitsIdentityError handling
01
MCP over CRMExpose account, opportunity and activity tools with typed schemas.
02
Safe write-backUpdate CRM idempotently with full attribution metadata.
03
Consent enforcementBlock an outbound action that consent does not permit.
04
API limitsHandle platform governor limits without losing work.
05
Error handlingEnsure a retry never double-sends a customer communication.
Assessment 06 — Systems integration
07Privacy, brand safety and securityConsent, privacy regimes and defending an agent that talks to the public.10 hrs · 5 labs
GDPR/CCPA obligationsConsent enforcementBrand safetyOWASP LLM Top 10Prompt injection from user inputEscalation policy
01
Consent gateEnforce consent and suppression at the tool layer and test it.
02
Data subject requestHandle deletion and access requests across agent memory.
03
Injection defenceDefend a customer-facing agent against adversarial user input.
04
Brand safetyPrevent off-brand or non-compliant output reaching a customer.
05
Red teamRun an OWASP LLM Top 10 pass on a public-facing agent.
Assessment 07 — Privacy, brand safety & security
08Production, measurement and costHoldouts, CRM health and the cost per interaction that decides whether it scales.10 hrs · 5 labs
Eval suitesHoldout designCRM quality monitoringDriftCost per interactionHandover
01
Eval gateGate deployment on resolution accuracy and compliance regression.
02
HoldoutDesign and analyse a holdout that credibly measures lift.
03
CRM healthProve the agent did not degrade record quality.
04
Cost per interactionMeasure and reduce cost per conversation.
05
HandoverProduce runbook, monitoring and owner assignment.
Assessment 08 — Production & operations
Two capstones — 20 hours
End-to-end agent design, development, deployment and testing, 10 hours each. Reviewed by a Senior SCDAI Engineer against a published rubric.
Capstone A — Inbound engagement and qualification agent
Build an agent that engages inbound interest instantly, qualifies conversationally, enforces consent and books a meeting — with a holdout proving incremental conversion.
Capstone B — Grounded service resolution agent
Build an agent that resolves cases from cited knowledge, executes the fix through an idempotent tool under approval, and demonstrably does not degrade CRM data quality.
What you deliver, and how it is marked
- A running agent, deployed on the program's target stack, reachable by a reviewer.
- An architecture note: boundaries, tools, autonomy level and the trade-offs you took.
- A versioned eval set with a baseline and your final scores.
- A trace walkthrough of one success and one deliberate failure.
- Cost per task, measured — not estimated.
- A runbook and rollback path a client team could operate on Monday.
Problem fit15%Is this a decision worth automating, scoped to a defensible boundary?
Architecture20%Right pattern for the constraints; autonomy earned, not assumed.
Grounding & accuracy20%Measured groundedness against the eval set, with the failures named.
Safety & compliance20%Domain controls actually enforced, and demonstrably so.
Operability15%Traces, runbook, rollback, cost — could someone else run this?
Communication10%Can a client executive follow the reasoning in ten minutes?
One lab, start to finish
Every lab in this program follows the shape below. This is lab 07 in full — the brief you are given, the environment that is provisioned for you, the code you start from and the assertions that decide whether you passed.
Module 07 · Privacy, brand safety and security
Consent enforcement at send time
Block outbound contact that consent does not permit, at the tool rather than in the prompt.
The brief
Build an outbound engagement agent over a seeded CRM. Consent changes during the run. The agent must check at send time, not at segment-build time, and suppression must override consent in every case.
Provisioned for you
- CRM: 20k contacts with consent history
- Consent registry with mid-run mutations
- Suppression list
- Send tool with audit capture
You start from this
def send(contact_id, channel, message, registry, suppression):
"""Consent is checked here, at the moment of action. A segment built
ten minutes ago is already stale."""
raise NotImplementedErrorWhat the grader asserts
- Zero sends to contacts whose consent was withdrawn during the run
- Suppression blocks the send even where consent is present
- A retry after timeout does not double-send
- Every send records channel, purpose, lawful basis and timestamp
- Segment-time consent is never trusted as send-time consent
05
Assessment
Assessed, not attended
Every module ends with a timed, randomised assessment delivered through the SCIKIQ Agentic AI Playground. The certificate requires a pass on all of them plus two reviewed capstones.
Assessment blueprint
What each test covers, how long it runs, and how many items are currently in the versioned bank behind it.
| # | Assessment & coverage | Items | Time | In bank |
|---|---|---|---|---|
| 01 | Operating model & scopingDelivery mandateJourney selectionValue sizingAdoption dynamics | 25 | 35 min | 4 |
| 02 | Commercial data landscapeCRM modelsIdentity resolutionConsentData quality | 25 | 35 min | 4 |
| 03 | Context engineeringBrand and claimsPersonalisation groundingEscalationCost | 25 | 35 min | 4 |
| 04 | Retrieval & groundingContent indexingCase retrievalFreshnessRetrieval metrics | 25 | 40 min | 4 |
| 05 | Agent design & autonomyConversational designHandoffCRM write disciplineContainment | 25 | 40 min | 4 |
| 06 | Systems integrationMCP over CRMWrite safetyConsent enforcementResilience | 25 | 35 min | 4 |
| 07 | Privacy, brand safety & securityConsent and privacyBrand safetyLLM securityEscalation | 30 | 45 min | 4 |
| 08 | Production & operationsEval gatingHoldout measurementData qualityEconomics and handover | 30 | 45 min | 4 |
Single best answerOne defensible option among plausible distractors drawn from real field mistakes.
Scenario judgementA client situation with constraints; you pick the action a Senior SCDAI Engineer would take.
Architecture selectionGiven non-functional requirements, choose the pattern and justify the trade-off.
Failure diagnosisA trace, an eval report or a cost curve — identify the root cause.
- Each module ends with a timed assessment; the program certificate requires a pass on every one.
- Items are drawn at random from a versioned bank, so no two learners sit an identical paper.
- Pass mark is 70%. Two retakes are included, each with a fresh draw.
- Scenario items carry double weight — they are the ones that predict field performance.
- Capstones are reviewed against a published rubric by a Senior SCDAI Engineer, not auto-graded.
Try four questions from the bank
Real items, drawn from 32 in this program's bank — weighted toward the scenario and diagnosis types, because those are the ones that predict field performance. Instant feedback, nothing saved.
Four sample items — one attempt each, then the reasoning is shown.
Q1Before promising personalisation, what must you verify?
Personalisation on a wrong identity is worse than generic messaging, and consent determines what is even permitted.
Q2Why is identity resolution the foundational layer here?
It is the layer nobody funds and everything depends on. Resolution quality is the ceiling on the whole deployment.
Q3Generated content makes a product claim not in the approved library. What control failed?
Public-facing claims can carry regulatory and contractual exposure. The gate must be pre-publication and hard.
Q4Why must approved content be version and market scoped?
Market and version scoping is a compliance requirement, not a content-management convenience.
06
Career & outcomes
What this makes you, and where it goes next
The same ladder whichever specialisation you enter through — what changes is the domain you go deep in. Below: how the program is delivered, the skills it moves, the roles it leads to, and the specialisations closest to this one.
Available two ways
The same labs, assessments and capstones, delivered to an enterprise cohort or to individual professionals.
B2B
Direct to enterprise
Cohorts of 20 to 2,000+ on your own tenancy, with your data patterns and your cloud. Skill-gap baselining up front, per-team mastery reporting throughout, and capstones scoped against your real backlog so the output is deployable work.
- Private cohort on your tenancy
- Baseline skills assessment before kick-off
- Capstones scoped to your backlog
- Per-team mastery dashboards for L&D
- Optional Senior SCDAI Engineer architecture reviews
B2C
Direct to individual professionals
The same labs, assessments and capstones for individual engineers and analysts, run on shared infrastructure with a fixed cohort calendar. You leave with a graded portfolio, not a certificate of attendance.
- Fixed-date cohorts, self-paced within them
- Same graded labs and item bank
- Portfolio-ready capstones you own
- Certificate on a pass in every module
- Community and office-hours access
Skill matrix
Find your row and aim one column right. The Playground scores you against this after every module.
| Skill | Beginner | Intermediate | Advanced |
|---|---|---|---|
| Commercial fluency | Knows the funnel. | Maps impact journeys to decisions and owners. | Sizes value and rewires a journey end to end. |
| Identity & consent | Aware of GDPR. | Enforces consent at the tool layer. | Designs identity resolution that personalisation can rely on. |
| Commercial measurement | Reports activity. | Designs holdouts for lift. | Defends incremental impact to a sceptical CFO. |
| Context engineering | Writes clear prompts. | Structures retrieval, tools and state deliberately. | Designs context strategy for reliability and cost at scale. |
| Retrieval & grounding | Builds basic vector search. | Tunes chunking, hybrid search and reranking. | Designs graph + vector grounding with measured recall. |
| Agent orchestration | Runs a single tool-calling agent. | Builds supervised multi-step and multi-agent flows. | Designs autonomy boundaries and failure containment. |
| Tool & system integration | Calls a documented API. | Writes an MCP server over a system of record. | Designs a least-privilege tool estate across systems. |
| Evaluation | Eyeballs outputs. | Builds labelled eval sets and regression gates. | Runs online evals with drift and judge calibration. |
| Observability & cost | Reads logs. | Traces runs, tracks tokens and latency. | Owns cost per task and capacity planning in production. |
| Security & guardrails | Adds output filters. | Mitigates the OWASP LLM Top 10 in a build. | Threat-models an agent estate and proves controls. |
| Client delivery | Takes notes in a workshop. | Runs discovery and scopes a thin slice. | Owns the account technically, from scope to handover. |
Where it takes you
The SCDAI ladder is the same whichever specialisation you enter through — what changes is the domain you go deep in.
01
Data and AI EngineerEntry
Python, LLM APIs, first RAG build, prompt and context basics.
02
Associate Data and AI EngineerPractitioner
Client-facing discovery, tool integration, agent build under supervision.
03
SCIKIQ Certified Data and AI EngineerSpecialist
Owns a deployment end to end — scoping, build, evals, handover.
04
Senior Data and AI Engineer / Domain LeadSenior
Multi-agent architecture, regulated deployments, reference patterns for the practice.
05
AI Architect / Practice LeadLeadership
Portfolio of accounts, delivery standards, hiring and capability strategy.
Terms worth knowing
- Impact journey
- An end-to-end commercial workflow rewired as a unit, rather than a tool added to a step.
- Speed to lead
- Time from inbound interest to first meaningful response — a dominant conversion factor.
- NRR
- Net Revenue Retention — retention plus expansion, the core growth metric.
- CPQ
- Configure, Price, Quote — the systems governing what can be sold and at what price.
- Suppression
- Rules preventing contact with specific individuals or segments.
- Holdout
- A withheld group used to measure incremental impact rather than attributed impact.
- Deflection
- Service demand resolved without a human agent.
- Agentic commerce
- Buying where an AI agent performs discovery and selection for the customer.
- SCDAI
- SCIKIQ Certified Data and AI Engineer — the credential this program awards.
- SCDAI engineer
- An engineer embedded with the customer who turns their constraints into a working system.
- Thin slice
- The narrowest end-to-end path through a workflow that still produces business value.
- Agentic RAG
- Retrieval where the agent decides when, what and how often to retrieve.
- MCP
- Model Context Protocol — open standard connecting an agent to tools and data.
- A2A
- Agent2Agent — protocol for agent-to-agent delegation across vendors.
- Eval set
- A labelled, versioned set of cases an agent must pass before release.
- LLM-as-judge
- Using a model to score outputs, calibrated periodically against human ratings.
- Guardrail
- An enforced constraint on an input, output or action — not a prompt request.
- Human-in-the-loop
- A required human approval step before a consequential action executes.
- Cost per task
- Total token, tool and compute cost of one completed unit of work — the engineer's unit economic.
- Golden path
- The reference architecture a practice standardises on so deployments stay reviewable.
- Handover
- The point at which the client can operate, evaluate and change the system without the engineer.
Related programs
Skill a team, or join a cohort
B2B cohorts run on your tenancy with capstones scoped to your backlog. B2C cohorts run on a fixed calendar.
07
Enrolling
What you need, what you get, and what it costs
Stated plainly enough to rule yourself in or out without a sales call: the prerequisites, how the program runs, exactly what the credential is worth, and the questions everyone asks.
Before you start
Stated plainly so you can rule yourself in or out without a sales call. Nothing here is a formal qualification — it is what the first lab assumes you can already do.
Technical
You should already be able to do these
- Comfortable in Python: functions, typing, virtual environments, reading a stack trace
- Have called an LLM API and handled a structured response
- Git, the command line, and reading someone else’s code without panic
Domain
What we assume, and what we teach
- Know what a CRM is and roughly what sits in one
- Comfort with the idea of consent and privacy obligations
Time and access
What the program asks of your week
- A machine that can run a browser and a terminal — labs run in the Playground, not locally
- Roughly 8 hours a week for the taught weeks, plus 20 hours for the two capstones
How it runs
Cohort dates and pricing are confirmed on enquiry rather than printed here, because both move with the intake.
- Format
- Self-paced within a fixed cohort window, delivered in the SCIKIQ Agentic AI Playground
- Taught hours
- 80 hours across 8 modules
- Capstones
- 20 hours — two end-to-end builds, reviewed not auto-graded
- Typical duration
- 10–12 weeks at around 8 hours a week
- Level
- Intermediate → Advanced
- Built for
- Engineers and analysts deploying AI across marketing, sales and service
- Next cohorts
- — confirmed on enquiry
- Pricing
- — confirmed on enquiry
The credential
The credential is awarded per specialisation, so it names the domain or stack you were assessed in rather than claiming general competence. On this program the badge reads SCDAI — Marketing, Sales & Service.
How it is earned
- A pass at 70% or above on every module assessment — not an average across them
- Both capstones reviewed against the published rubric by a Senior SCDAI Engineer
- All labs submitted and machine-graded green
- Awarded as
- A verifiable digital badge with a public credential record
- Valid for
- Two years from award
- Recertification
- One current-year assessment plus a refreshed capstone
- Retakes
- Two included per assessment, each with a fresh item draw
- Named specialisation
- The badge states the domain or stack, e.g. SCDAI — Healthcare
- Verification
- Third parties confirm a holder from the credential record, no login required
A certificate that cannot be checked is decoration. Every award resolves to a record showing the specialisation, the award date and the assessments passed.
The certificate you leave with
This is the credential the program awards, shown exactly as it is issued — with the specialisation named, the assessment record attached and a verification link anyone can check without an account.
SCIKIQ Data Academy
Certificate of Professional Certification
This is to certify that
Your name
has been assessed and certified as
SCIKIQ Certified Data and AI Engineer
Marketing, Sales & Service
- Specialisation
- Marketing, Sales & Service
- Assessments passed
- 8 of 8, each at 70% or above
- Labs graded
- 40 of 40 machine-graded green
- Capstones reviewed
- 2, against the published rubric
- Taught hours
- 80 across 8 modules, plus 20 capstone hours
AwardedOn completion
Valid untilTwo years from award
Credential IDSCDAI-MARKETING-SA-<issued>
Senior SCDAI Engineer, Certification Board
- 01The specialisation is on the certificate
Not “Data and AI Engineer” but the domain or stack you were actually assessed in. A general claim would be a weaker one.
- 02The assessment record travels with it
Modules passed, labs graded and both capstones reviewed — so the credential states what was measured rather than that you attended.
- 03Anyone can verify it, without a login
The credential ID resolves to a public record showing the specialisation, the award date and the assessments passed.
Share it on LinkedIn
Add it to your LinkedIn profile in one step. The link pre-fills the certification fields from the credential record, so the entry on your profile matches the record a reader can check.
Add to LinkedIn profile The button is live on your real certificate; here it opens LinkedIn pre-filled with this specialisation so you can see exactly what the profile entry will say.A certificate that cannot be checked is decoration. Every award resolves to a record showing the specialisation, the award date and the assessments passed.
Questions we get asked
The objections that come up in every conversation about this program, answered without the brochure voice.
Is this a course or a certification?
Both, and the second is the point. Eight timed assessments and two reviewed capstones stand between you and the credential, so a pass means someone measured the skill rather than recorded your attendance.
How is it self-paced without falling apart?
Every lab is provisioned, graded and unblocked by the Playground rather than by an instructor. That is what lets a cohort of 2,000 cost the same faculty time as a cohort of 20 — and why you are never waiting on someone to mark your work.
What if I fail an assessment?
Two retakes are included per assessment, each drawing a fresh item set from the bank, so a retake is a genuinely new paper rather than the same questions again.
Do I need my own cloud account?
For the domain specialisations, no — labs run in provisioned sandboxes. For the four tech-stack programs you will want access to that platform, since deploying into a real subscription is much of the point.
Can we run this privately for our team?
Yes. B2B cohorts run on your own tenancy with your data patterns, a skills baseline before kick-off, per-team mastery reporting, and capstones scoped against your actual backlog so the output is deployable work rather than an exercise.
Which specialisation should I take?
Take the domain you deploy into. If you move across industries, take a tech-stack program instead and pick up domain context on the engagement. The chooser on the programs page will narrow it.
How current is the material?
The trends, platform capabilities and regulatory positions are reviewed each quarter, and every external claim on these pages links to its source so you can check the date yourself.
What do I actually leave with?
A graded portfolio: forty machine-graded labs, two reviewed end-to-end agent builds with measured evaluation and cost per task, and a verifiable credential naming your specialisation.
Still deciding?
Tell us the systems you deploy into and we will say plainly whether this specialisation is the right one — or which of the 21 is.