Skip to content

Online + offline · Bengaluru

Generative AI Course: Become a GenAI / RAG Engineer

Become a GenAI engineer who builds RAG systems: embeddings, vector databases, LangChain and evaluation. Live batches in Bangalore, placement support.

  • 15+ placed every month
  • Support until you are placed

The Generative AI track teaches working professionals, starting from Python and SQL, to build retrieval-augmented generation (RAG) systems that hold up outside a demo: chunking, embeddings, vector databases, retrieval and re-ranking, LangChain and LlamaIndex orchestration, evaluation, guardrails and deployment. It runs three months alongside a full-time job, with unlimited mock interviews and placement support until you are hired.

Who this track is for

Working professionals with two or more years of experience, IT or non-IT, and people returning after a career gap.

  • IT professionals: Developers and data engineers who want to build RAG and LLM systems rather than read about them.
  • Non-IT professionals: Analysts and domain experts who want to move into applied AI work.
  • Returning after a career gap: Professionals coming back after a break who want the most in-demand skill set and help explaining the gap.

What you will be able to do

Own the work end to end: design it, build it, explain it in an interview, and keep it running once it is live. Every module ends with something you can show a hiring manager rather than a certificate.

Syllabus

What you learn, module by module

7 modules and 28 topics over 3 months, live, alongside your job.

  1. Python for AI engineering 4 topics · Core Python: functions, classes, error handling and virtual environments, …
    • Core Python: functions, classes, error handling and virtual environments
    • Working with files, JSON and REST APIs; async calls with asyncio
    • pandas for preparing documents, tables and evaluation data
    • Pydantic models for structured LLM output
  2. SQL and data for AI apps 4 topics · SQL essentials: joins, GROUP BY, CTEs and window functions, …
    • SQL essentials: joins, GROUP BY, CTEs and window functions
    • PostgreSQL with pgvector for storing embeddings next to your data
    • Text-to-SQL: letting an LLM answer questions from a database safely
    • Logging questions, answers and feedback for evaluation
  3. LLM fundamentals 4 topics · Tokens, context windows and how LLMs use retrieved context, …
    • Tokens, context windows and how LLMs use retrieved context
    • Prompting for grounded, cited answers
    • Calling LLM APIs: streaming, tool calling and retries
    • Cost, latency and model selection
  4. Ingestion, chunking and embeddings 4 topics · Loading PDFs, web pages and tables cleanly, …
    • Loading PDFs, web pages and tables cleanly
    • Chunking strategies and chunk-size trade-offs
    • Choosing and comparing embedding models
    • Metadata, filtering and keeping the index up to date
  5. Vector databases and retrieval 4 topics · Vector indexes and similarity search, …
    • Vector indexes and similarity search
    • Hybrid search: keyword plus vector
    • Re-ranking retrieved chunks with cross-encoders
    • Query rewriting and multi-query retrieval
  6. Orchestration with LangChain and LlamaIndex 4 topics · RAG chains, retrievers and query engines, …
    • RAG chains, retrievers and query engines
    • Conversation memory and source citations
    • Tool calling and agentic RAG, and when a plain pipeline is better
    • Streaming and caching for user-facing latency
  7. Evaluation, guardrails and deployment 4 topics · Building an evaluation set from real questions, …
    • Building an evaluation set from real questions
    • Measuring retrieval recall, faithfulness and answer relevance
    • Guardrails: prompt injection, sensitive data and refusals
    • Deploying with FastAPI, monitoring and cost tracking

Portfolio

3 projects you can explain in interviews

Built on real company-style problems, and written up so you can walk an interviewer through them.

  1. RAG over real documents

    Ingest a messy document set, then tune chunking, embeddings and re-ranking, measuring how each change moves answer quality.

  2. RAG evaluation harness

    Build a test set and scoring pipeline for retrieval recall and faithfulness that shows whether a change made answers better or worse.

  3. Production RAG service

    Ship a cited-answer assistant behind a FastAPI endpoint with guardrails, streaming, caching and cost tracking.

Mock interviews

Practice until real interviews feel easy

Every mock looks like the real thing, and there is no limit on how many you get.

  1. Start early

    Your first mock comes before you feel ready. It shows the gap between what you know and what you can say under pressure.

  2. Weekly and recorded

    A mock with the placement team every week, each one recorded and followed by written feedback on what to fix.

  3. Real company rounds

    Technical, system design and manager rounds, modelled on the companies you are actually targeting.

  4. No cap

    Mocks continue until you are clearing real interviews, not until a quota runs out.

Proof

Placed from this track

All placement stories →
  1. Before Data Analyst Cognizant
  2. After GenAI Engineer Fractal

Irfan worked in reporting and dashboards, and wanted to move into applied AI before that window closed. He built two retrieval-augmented generation systems and, more unusually, an evaluation harness that measured answer quality. He joined Fractal as a GenAI Engineer three months after his first assessment call.

Read Mohammed's story

Mohammed's move

Time to the new role
3months
Salary, per year
₹11L → ₹24L

Across Stint Academy

Placed every month
15+
Working professionals, not freshers
Hiring companies
100+
MNCs, startups and product companies

Mohammed's numbers are from their own placement. Results depend on your experience and interviews.

Fee

Most of your fee is paid after you're placed

  • Training + placement ₹3,50,000 incl. GST
  • Placement only ₹3,20,000 incl. GST

Not placed within 3 months? We return your ₹50,000 token. The ₹2,70,000 balance is due only if you're placed.

How the payments split

  1. Training fee ₹30,000 At enrolment · Placement only: ₹0
  2. Placement token ₹50,000 At enrolment · Refundable
  3. Placement fee balance ₹2,70,000 After you're placed · Only when placed

Both plans include GST. Pay only into the account of TMG Edutech Pvt Ltd and keep every receipt.

FAQ

Questions about this track

Straight answers to what people ask before they book a call.

Can I do this track while working full time?

Yes. Batches run on weekday evenings and weekends, sessions are recorded, and mock interviews are scheduled around your working hours.

What do I need to know before starting the Generative AI track?

Two or more years of work experience and comfort with basic logic. We cover the technical foundations from the start, so you do not need prior experience with these specific tools.

What happens if I do not clear interviews in the first few attempts?

Mock interviews continue without a cap and placement support runs until you are hired. Most people need several attempts, and the feedback after each one is the point.

Free assessment

Your next role starts with one call.

Not sure which track fits you? Book a free 30-minute assessment.

  • Talk to a real counselloryour target role and timeline

  • Get a recommended trackmatched to you, not your title

  • Know the feemost of it is paid after you’re placed

Free · 30-minute call

Book your free assessment

Tell us where you are today and we'll map the right track for you.