🚀 TECH 301 – SENIOR PRACTITIONER
Program Title: AI Systems Engineering
Duration: 14 Sessions (2 Hours Each)
🎯 Program Objective
The objective of TECH 301 is to enable senior engineers to design, evaluate, and deploy AI-enabled systems that are scalable, reliable, cost-optimized, and production-ready.
Participants will learn how to build AI systems using architectures such as Retrieval-Augmented Generation (RAG), implement observability mechanisms, define evaluation metrics, and introduce guardrails for safe deployment.
Participants will learn how to build AI systems using architectures such as Retrieval-Augmented Generation (RAG), implement observability mechanisms, define evaluation metrics, and introduce guardrails for safe deployment.
✅ Pre-Requisites
- 3–6 years of development experience
- Strong understanding of system design and architecture principles
- Experience building and deploying backend services
- Familiarity with APIs and cloud-based services
- Basic understanding of REST, microservices, and distributed systems
📘 Curriculum Overview
| Session | Topic | Detailed Coverage | Tools | Learning Outcome |
|---|---|---|---|---|
| 1 | AI vs Rules | Evaluate deterministic logic vs AI approaches. | ChatGPT | Choose appropriate system approach. |
| 2 | AI Trade-offs | Analyze cost, latency & accuracy trade-offs. | Azure OpenAI | Make informed architecture decisions. |
| 3 | Prompt Systems | Design structured, versioned prompt frameworks. | GitHub | Create reusable prompt architectures. |
| 4 | Prompt Evaluation | Test outputs for consistency & reliability. | ChatGPT | Improve output quality via evaluation. |
| 5 | Embeddings | Vector representations & semantic search foundations. | OpenAI | Understand embedding-based retrieval. |
| 6 | Retrieval Systems | Build context retrieval pipelines. | FAISS / Pinecone | Design intelligent retrieval systems. |
| 7 | RAG Architecture | Combine retrieval + generation. | LangChain | Build context-aware AI systems. |
| 8 | RAG Optimization | Improve chunking, ranking & filtering strategies. | LangChain | Optimize system accuracy. |
| 9 | AI Observability | Track logs, outputs & performance metrics. | LangSmith | Monitor AI behavior effectively. |
| 10 | Feedback Loops | Design human-in-the-loop improvement systems. | ChatGPT | Implement continuous improvement. |
| 11 | Failure Modes | Analyze hallucinations & reasoning failures. | ChatGPT | Mitigate AI risks. |
| 12 | Guardrails | Implement output constraints & filters. | Azure AI | Enforce safe AI behavior. |
| 13 | Scaling AI | Design scalable AI infrastructure. | Azure OpenAI | Scale AI workloads efficiently. |
| 14 | Cost Optimization | Reduce token usage & optimize operations. | Azure | Manage AI cost effectively. |
| 15 | Production Deployment | Deploy AI with monitoring & fallback strategies. | Azure | Deploy reliable AI services. |
| 16 | Case Study | Build production-grade RAG solving real problem. | All Tools | Design & deploy complete AI system. |