SCIKIQ Certified
Data and AI Engineer
21 domain and tech-stack specialisations that turn engineers into Certified Data and AI Engineers — the people who take an AI system into a client’s messy reality and make it work. Designed and delivered by Senior SCDAI Engineers with a decade of global AI implementation and agentic product engineering behind them.
Why these programs are different
A generic agentic AI course teaches you to build an agent. These teach you to deploy one inside a bank, a hospital, a refinery or a plant — where the wall you hit is different every time.
Domain-specific, not generic
A banking engineer and a life-sciences engineer hit different walls. The labs, the data and the assessments are built for the wall you will actually hit.
Built by practising Senior SCDAI Engineers
Designed and delivered by engineers with a decade of global AI implementation and agentic product work behind them — the patterns come from deployments, not slides.
Scales without instructors
Every lab is provisioned and graded by the Playground, so a cohort of 2,000 costs the same faculty time as a cohort of 20.
Assessed, not attended
Eight timed assessments and two reviewed capstones. The certificate means someone measured the skill.
Which one is right for you
21 programs is a wall. Two questions narrow it to the one or two that actually apply — what you deploy into, and what you build on.
Pick a sector or function above.
Pick a platform above.
Take the domain program first — it is the one that carries the credential you will be hired against. The stack program then costs you 60 hours and makes you portable across engagements.
The domain program stands on its own; the labs run in provisioned sandboxes, so no cloud account of your own is needed. Add a stack program once you know what you will be deploying on.
We do not have a specialisation for your sector yet, and routing you to the nearest-looking one would not serve you. Take the stack program instead and pick up domain context on the engagement — that is what most consultants do.
Answer both questions and we will narrow the set to two. If neither fits, the comparison below shows all of them side by side.
The 21 programs
17 domain specialisations at 80 hours with 40 labs and 8 assessments; 4 tech-stack specialisations at 60 hours with 20 labs and 4 assessments. Every one adds two 10-hour end-to-end capstones.
Financial services
Regulated money: lending, markets, cover and the rails between them.
4Banking & Financial Services
For bankingThe value chain, the data and architecture, and the agentic use cases that survive a regulator.
Investment Banking
For advisory and marketsThe origination-to-settlement value chain, deal and market data architecture, and agentic use cases built around the information barrier rather than despite it.
Cards & Payments
For issuers, acquirers and processorsThe authorisation-to-settlement value chain, PCI-scoped data architecture, and agentic use cases on a path measured in milliseconds.
Insurance
For insuranceThe value chain from product to claim, the data and architecture, and the agentic use cases carriers are scaling.
Health & life sciences
Care delivery and the products behind it, where privacy and validation bind.
2Healthcare
For healthcareThe care and revenue value chain, FHIR-era architecture, and agentic use cases that survive HIPAA.
Life Sciences / Pharma
For pharma and life sciencesThe R&D-to-commercial value chain, GxP-validated architecture, and agentic use cases regulators will accept.
Consumer & commerce
Making it, selling it and shipping it — three businesses, not one.
3FMCG & CPG
For brand ownersThe innovation-to-deduction value chain, master data and syndicated data architecture, and agentic use cases where you never observe the shopper directly.
Retail
For retailersThe range-to-returns value chain, store and inventory data architecture, and agentic use cases where the constraint is what is physically on a shelf.
Ecommerce & Marketplaces
For online retail and marketplacesThe traffic-to-retention value chain, catalogue and behavioural data architecture, and agentic use cases on a latency budget the shopper can feel.
Industry, energy & the built environment
Physical assets with long lives, unforgiving failure modes and statutory duties.
3Oil & Gas
For energyThe upstream-to-downstream value chain, industrial data architecture, and agentic use cases where safety is the binding constraint.
Manufacturing & Supply Chain
For industrial operationsThe plan-source-make-deliver chain, ERP and shop-floor architecture, and agentic use cases that reach execution.
Facility Management
For built-environment operationsThe portfolio-to-occupant value chain, building systems and CAFM data architecture, and agentic use cases where statutory compliance sets a floor nothing may fall below.
Networks & transport
Infrastructure businesses judged on uptime, recovery and what the customer felt.
3Telecom
For operatorsThe network-to-customer value chain, OSS/BSS data architecture, and agentic use cases where the network event and the customer experience are the same fact seen twice.
Telecom Towers
For passive infrastructureThe site acquisition-to-tenancy value chain, energy and uptime data architecture, and agentic use cases where the customer is an operator with an SLA and a penalty clause.
Airlines
For carriersThe schedule-to-settlement value chain, PSS and operations data architecture, and agentic use cases where the binding constraint is a network that either recovers or does not.
Cross-industry functions
The functions that exist in every sector, and travel with you between them.
2Finance, Accounting & Auditing
For the finance functionThe record-to-report value chain, controls-aware architecture, and agentic use cases that pass a SOX auditor.
Marketing, Sales & Service
For commercial functionsThe revenue value chain, CRM and data architecture, and agentic use cases that actually convert.
Tech stacks
Platform depth rather than sector depth — take one when you move across industries.
4OpenAI + Azure Tech Stack
For the microsoft stackFoundry, Agent Service, grounding, evaluation and governed enterprise deployment.
Anthropic + AWS Tech Stack
For claude and awsBedrock AgentCore, long-running agents, MCP tooling, evaluation and production operations.
Google (GCP, Gemini & Gemma, ADK)
For google cloudADK, Agent Engine, Gemini and Gemma, grounding on BigQuery, evaluation and production operations.
Open Source & Open-Weight LLMs / SLMs
For self-hosted modelsServing with vLLM, fine-tuning, evaluation, GPU economics and sovereign deployment.
All 21, side by side
Same structure, same pass mark, same two capstones on every one. What changes is the domain or stack you go deep in, and how many taught hours that takes.
| Program | Type | Taught hrs | Labs | Assessments | Capstone hrs | Built for |
|---|---|---|---|---|---|---|
| 01Banking & Financial Services | Domain | 80 | 40 | 8 | +20 | Engineers, architects and analysts deploying AI inside banks |
| 02Investment Banking | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across advisory, research and markets |
| 03Cards & Payments | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across issuing, acquiring and payment processing |
| 04Insurance | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI at carriers, brokers and MGAs |
| 05Healthcare | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI at providers, payers and health tech |
| 06Life Sciences / Pharma | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI in GxP and commercial pharma environments |
| 07FMCG & CPG | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI at consumer goods manufacturers |
| 08Retail | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across retail merchandising and store operations |
| 09Ecommerce & Marketplaces | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across ecommerce and marketplace operations |
| 10Oil & Gas | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across energy operations |
| 11Manufacturing & Supply Chain | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across manufacturing and supply chain |
| 12Facility Management | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across facilities and building operations |
| 13Telecom | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI inside network and customer operations |
| 14Telecom Towers | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across passive telecom infrastructure |
| 15Airlines | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across airline commercial and operations |
| 16Finance, Accounting & Auditing | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI in finance, accounting and audit functions |
| 17Marketing, Sales & Service | Domain | 80 | 40 | 8 | +20 | Engineers and analysts deploying AI across marketing, sales and service |
| 18OpenAI + Azure Tech Stack | Tech stack | 60 | 20 | 4 | +20 | Engineers deploying agents on Microsoft Foundry and Azure |
| 19Anthropic + AWS Tech Stack | Tech stack | 60 | 20 | 4 | +20 | Engineers deploying Claude-based agents on AWS |
| 20Google (GCP, Gemini & Gemma, ADK) | Tech stack | 60 | 20 | 4 | +20 | Engineers deploying agents on Google Cloud |
| 21Open Source & Open-Weight LLMs / SLMs | Tech stack | 60 | 20 | 4 | +20 | Engineers deploying self-hosted open-weight models and agents |
How every program is built
The same three-part structure, whichever specialisation you take.
Part 1 — The business
The value chain end to end, the personas whose week changes, and what McKinsey, BCG, Bain and HBR are seeing in that sector right now.
Part 2 — The technology
The data landscape you will actually meet, the reference architecture that gets approved, and the constraints that shape it.
Part 3 — Getting it done
Use cases that join a value-chain stage to an architecture layer, the delivery playbook, 760 labs and two capstones.
Then — Assessment
152 timed assessments across the portfolio, drawn from a versioned item bank, plus capstones reviewed against a published rubric.
The SCDAI delivery model
The through-line of every program, taught in module 01 and assessed in every scenario item.
Find the decision a human makes today, how often, and what it costs when it is wrong. That is the agent boundary.
A working thin slice on the client's own messy data beats a clean demo on yours. Ship it in the first week.
Start as a workflow with human approval. Widen the agent's authority only where evals prove it has earned it.
Traces, evals and cost per task from day one — otherwise you cannot tell improvement from drift.
Runbook, eval suite, rollback path and a named owner. The exit criterion is that they can change it without you.
Built to scale without instructors
The lab and assessment platform behind every program. It provisions the environment, seeds the data, runs the graders and keeps the audit trail — which is what lets one faculty team skill thousands of engineers at once.
Self-paced, zero instructor dependency
Every lab is provisioned, graded and unblocked by the Playground itself. Scaling from 20 learners to 2,000 costs no additional faculty time.
Ephemeral, pre-seeded lab environments
Each lab spins up its own sandbox with realistic seeded data, the model endpoints already wired and the reference solution held back until you submit.
Machine-graded, not opinion-graded
Labs are scored by assertion suites and eval harnesses that run against your agent, so the same standard applies to every learner in every cohort.
Senior SCDAI Engineer office hours, on the exceptions
Faculty time goes where automation cannot: architecture reviews, capstone critique and the escalations the Playground flags.
The SCIKIQ Agentic AI Playground
The lab and assessment platform behind every program. It provisions the environment, seeds the data, runs the graders and keeps the audit trail — which is what lets one faculty team skill thousands of engineers at once.
Provisioned sandboxes
Model endpoints, vector store, sample corpus and system-of-record stubs, ready on open.
Assertion-based lab grading
Your agent is executed against held-out cases; the grader scores behaviour, not prose.
Eval harness built in
Groundedness, task success, tool-choice accuracy and cost per task, tracked run over run.
Timed, randomised assessments
Each attempt draws a fresh item set from the bank, so a retake is a new test.
Trace-level review
Every agent run is traced with OpenTelemetry GenAI conventions — learners debug from spans.
Cohort analytics
Per-skill mastery across the org, so L&D sees where a team is thin before a project does.
Available two ways
The same labs, assessments and capstones, delivered to an enterprise cohort or to individual professionals.
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
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 20 engineers, or 2,000
Because the Playground provisions and grades every lab, a cohort of 2,000 costs the same faculty time as a cohort of 20.