Enterprise Generative AI & Agent Engineering (Azure)
Build production-minded GenAI systems: Azure OpenAI and AI Foundry, retrieval-augmented generation, LangChain and LangGraph, Semantic Kernel, and multi-agent workflows - with responsible AI, evaluation, and enterprise patterns. Designed for technical staff who already ship software, not for a long theory-only fundamentals track.
Certification exams are optional. Curriculum topics overlap with AI-901, AI-103, and AB-100 (agentic and enterprise AI solution design) where they support real implementation skills.
Who, when, how
Live sessions are paced for full-time work: concentrated weekends, async labs, and a capstone you can align with your own use cases (without using employer confidential data in class artifacts).
Audience
- Software / platform engineers and technical leads
- Teams modernizing on Azure and enterprise AI
- Anyone comfortable coding who needs GenAI + agents depth
Schedule
- 8 weekends · ~4-5 hours live per session
- Async lab exercises and readings between sessions
- Pilot cohort: 10-15 seats
Deliverables
- Session recordings and shared lab materials
- Guided capstone with instructor review
- Certificate of completion
Week-by-week schedule
Eight weekend modules build toward a capstone. Each week lists live focus, hands-on lab, and a capstone milestone so progress compounds.
Week 1 - Enterprise GenAI foundations
Weekend moduleLLMs at a working level: tokens, context windows, embeddings overview. Limitations, hallucinations, bias. Responsible AI and what regulated environments care about.
Baseline prompting patterns; compare outputs across settings.
Choose capstone problem statement (internal-style scenario, synthetic or public data only).
Week 2 - Azure OpenAI & AI Foundry
Weekend moduleModel deployment patterns, quotas, keys vs managed identity at a high level, private endpoints / RBAC concepts, cost awareness.
Call Azure OpenAI from a small app or notebook; manage configuration safely.
Define data sources and non-goals for your capstone.
Week 3 - Chunking & embeddings pipeline
Weekend moduleText chunking strategies, embedding models, batching, and ingestion hygiene.
Build an ingestion pipeline for a document set (sample corpus provided).
Lock corpus and chunking approach for capstone.
Week 4 - RAG with Azure AI Search
Weekend moduleVector + hybrid search, indexing patterns, grounding answers in sources.
End-to-end RAG MVP on Azure AI Search (or equivalent pattern).
Working RAG query path for capstone documents.
Week 5 - Evaluation & guardrails
Weekend moduleRAG evaluation basics (manual rubrics, simple automated checks). Hallucination mitigation. Prompt injection awareness and mitigations at a practitioner level.
Add a minimal evaluation harness or checklist to your RAG flow.
First evaluation results documented for capstone.
Week 6 - LangChain & LangGraph
Weekend moduleChains vs graphs; tool calling; durable multi-step flows; when graphs beat ad-hoc scripts.
Implement a short LangGraph workflow with tools (sample tools OK).
Integrate tool-calling or a small graph into capstone if fit.
Week 7 - Semantic Kernel & multi-agent exposure
Weekend moduleSemantic Kernel plugins and planners (Azure-aligned). Multi-agent patterns. Brief exposure to AutoGen or comparable framework as time allows - concepts over framework tourism.
SK plugin + planner exercise; sketch a two-role agent flow.
Optional: multi-step agentic path for capstone.
Week 8 - Production patterns & capstone demo
Weekend moduleObservability concepts, logging/tracing, API boundaries, how GenAI sits next to existing services. Banking-adjacent patterns (policy-grounded assistants, document workflows) as examples only.
Hardening checklist for your project; prepare demo.
Present capstone; written summary and repo handoff.
Optional add-ons when time permits: Prompt Flow / low-code orchestration touch, Copilot Studio overview, GitHub Copilot for productivity - labeled clearly so core depth is not diluted.
Pilot cohort
Individual enrollment. Timing and next steps are shared with people on the waitlist when a cohort date is set.
What's included
- 8 live weekend sessions (recordings provided)
- Hands-on labs and sample repositories
- Capstone review and demo feedback
- Certificate of completion
Cohort dates & waitlist
Next start date: to be announced to the waitlist. For fastest replies, use the email button below with your LinkedIn and a one-line background.
Practical details for accepted participants (including any cancellation terms) are provided directly by the organizer - they are not listed on this page.
Request a seat / waitlist