Applied AI Engineer Program. 150 hours. Two tracks.
LangChain · LangGraph · RAG · Production Agents · Evaluations · MCP. One course, two delivery modes — self-paced for working professionals (with 3 live mentor sessions/week) or full-time daily classes for freshers (~5 months).
Practitioner-taught · 25-student cohorts · Activation-code portal access
Pricing
₹50,000 course. ₹35,000 with masterclass coupon.
Same price for both tracks. Attend any free 2-hour masterclass to get the ₹15,000 coupon. Same final cost whether you’re a working pro or a fresher.
Course fee
₹50,000
150 hours of structured content + portal access (12 months) + cohort live mentor sessions. Same fee for the working-pro and fresher tracks.
With masterclass coupon
₹35,000
Attend any free 2-hour masterclass and you receive a ₹15,000 coupon on the call. Mention it when you contact us to enrol — we apply it before sending the Razorpay link.
How enrolment works
- 1. Contact us on WhatsApp or attend a free masterclass first.
- 2. We answer your questions on a quick call and confirm the right track for you.
- 3. We send a Razorpay payment link (with your masterclass coupon applied, if any).
- 4. After payment we email you an activation code.
- 5. Sign in to the Nettms portal, enter the code, and the full course unlocks instantly.
Why this course
Not another survey course. A working engineer’s curriculum.
Four things make this course different from the dozens of free Gen AI tutorials, the giant MOOCs, and the survey-level upskilling programs.
Production-grade, not survey-level
Every module ends in a working artifact you can run, not a quiz. You finish the course with a public GitHub showing real RAG pipelines, agents, and evals — the things hiring panels actually probe for.
Practitioner-taught, small cohort
Capped at 25 students. Instructors ship the same stack at paying clients every week — not academics, not content-only YouTubers. You learn the decisions a senior engineer actually makes.
Built for how you actually live
Working pros: pre-recorded portal videos + 3 live mentor sessions/week, so you can keep your job. Freshers: structured 3–4 hrs/day cohort with offline / online / hybrid options.
Real evaluations, not Andrew-Ng vibes
Every capstone is reviewed by a working engineer. You get specific feedback on prompt design, retrieval quality, eval coverage, and cost — not just "great job, here's a certificate".
What you’ll build
6 real systems. Yours to ship.
Every module produces a working artifact you can put on GitHub. By the end of 150 hours, you have a public portfolio that proves senior-level Gen AI engineering — not a certificate, real code.
- Capstone-eligible
Production RAG over your own documents
Chunking strategies, embedding choice, hybrid search, reranking, citation-grounded answers. The 2026-standard "chat with your docs" pipeline at production quality.
- Capstone-eligible
Multi-agent workflow with LangGraph
A stateful agent system that plans, calls tools, persists memory across sessions, and handles human-in-the-loop checkpoints. The senior-level pattern most Indian Gen AI roles screen for.
- Working pro favourite
Automate a real workflow with MCP
Build an MCP server exposing internal tools to Claude / Cursor. Working pros bring a real workflow from their job; we make the LLM drive it end-to-end without manual copy-paste.
- Advanced module
Fine-tune a Llama with LoRA
Take an open-weight model, fine-tune on a domain dataset (you choose: legal, support, e-commerce), deploy quantized to a single GPU. Cuts inference cost 80% vs Claude/GPT.
- Every project ships with this
Eval + guardrail suite
Golden dataset, LLM-as-judge, regression tests, prompt-injection defence, content moderation. The discipline that separates engineers from prompt tinkerers.
- Senior-engineer differentiator
Cost-optimised inference pipeline
Prompt caching, model routing (Haiku vs Sonnet vs Opus), batching, speculative decoding. Cut your API bill 60-80% without losing quality — the unsexy skill that promotes you.
Two tracks · same course
Pick the delivery mode that fits your life.
Same 12 modules. Same mentors. Same capstone. The only difference is pace and format.
For Working Professionals
Self-paced + 3 live mentor sessions/week
150 hours · self-paced delivery · ~12–16 weeks at evening pace
- All 12 modules as pre-recorded sessions inside the Nettms portal — watch on your schedule.
- 3 live mentor sessions per week (Tue/Thu/Sat evenings IST) to unblock you on real production work.
- Continue your current job. The course is built to run alongside, not interrupt.
- Bring real problems from your job to the capstone — turn cohort work into something you can ship at the office.
Best for: Software engineers with 2+ years of production experience who want to specialise into LLM / Gen AI engineering without quitting their job.
For Freshers
3–4 hrs/day · ~5 months · offline / online / hybrid
150 hours · structured cohort · ~5 months at 3–4 hrs/day
- Daily 3–4 hour classes (live + recorded) for ~5 months. Built to match a fresher’s available time.
- Choose the format that fits — offline at our Hyderabad lab, online from anywhere, or hybrid.
- Same 12 modules and same mentors as the working-pro track; just delivered at a slower, more structured pace.
- Capstone is a portfolio brief — ship a public Gen AI project you can put on a resume + GitHub.
Best for: B.Tech / BCA / B.Sc / MCA graduates (2024–26 batches) who want a structured runway into AI engineering and are willing to commit 3–4 hours daily for ~5 months.
Curriculum
12 modules. 150 hours of structured depth.
The senior-level Gen AI engineering stack — what Indian product companies actually probe for in 2026 interviews. No survey content; every module ends in a hands-on artifact.
- 1
Python for AI Engineering
Modern Python (3.12+), async, typing, project layout. The version of Python a working AI engineer uses in production.
- 2
LLM Fundamentals
How transformers, tokens, context windows, and inference economics actually work. Choosing between GPT, Claude, Gemini per task.
- 3
Prompt Engineering at Production Quality
RCTF framework, chain-of-thought, structured outputs, few-shot patterns. Designing prompts that hold up to 10,000 calls.
- 4
Retrieval-Augmented Generation (RAG)
Chunking strategies, embedding models, vector databases (pgvector + Pinecone), reranking. Build a real "chat with your docs" product end-to-end.
- 5
Function Calling, Tool Use & Structured Output
JSON schema, OpenAI tools, Claude tool use. Make the LLM do real work, not just generate text.
- 6
Agentic AI — The ReAct Loop
Plan → act → observe loops. The mental model behind every modern agent — Cursor, Claude Code, customer-support bots.
- 7
LangChain & LangGraph in Production
Stateful agent workflows. Human-in-the-loop checkpoints. Persistent memory. The framework most Indian product companies build on.
- 8
Multi-Agent Systems
Supervisor/worker, debate, pipeline patterns. When to use multi-agent (and when one well-prompted agent wins).
- 9
Evaluation & Guardrails
Golden datasets, LLM-as-judge, regression tests, content moderation, prompt-injection defence. The discipline that separates engineers from prompt tinkerers.
- 10
Cost & Latency Engineering
Prompt caching, model routing, quantization basics, batching. Cut your API bill by 60–80% without losing quality.
- 11
MCP (Model Context Protocol) & Agentic IDEs
Build MCP servers exposing your tools to Claude / Cursor. Drive Claude Code effectively. The 2026 baseline for AI engineering velocity.
- 12
Capstone — Ship a Production Agent
Bring a real workflow from your job (working pros) or a portfolio brief (freshers). Build, evaluate, deploy. Walk out with a public GitHub + a written architecture doc.
Stack
What you’ll be fluent in.
The exact tools the senior engineers at Razorpay, CRED, Fractal, Tredence and the rest of the Indian Gen AI hiring panel ship in production.
- LangChain
- LangGraph
- RAG + pgvector
- OpenAI / Claude APIs
- Python 3.12 + async
- MCP servers
- Evaluation + Guardrails
- Prompt caching
- Production agents
Lead instructor
Taught by K Shiva. A practitioner, not a YouTuber.
Shiva runs the IT/AI division at Nettms. He ships Gen AI systems for paying clients every week — production RAG pipelines, multi-agent workflows, MCP integrations, the exact stack you’ll learn. The course is taught from real engineering decisions, not generic tutorials.
- Builds Gen AI systems for paying clients across India and the Gulf.
- Personally reviews every cohort capstone — you get senior-engineer feedback, not auto-graded quizzes.
- Available in the live mentor sessions every week — no "assigned TA" middle layer.
- No co-branded rented credentials. The course stands on the work it produces.
Hyderabad · Nettms Urban Habitat IT/AI division
A typical week
What your week actually looks like.
No surprises. Here’s the real cadence for both tracks — what time you spend live, what you watch on your own, where mentors show up.
Working Professional · ~10 hrs/week
Mon–Wed
2 hrs/evening · pre-recorded module videos in the portal · code along at your own pace
Tue (7–8pm IST)
Live mentor session · bring a real production blocker or a module question · 1:1 priority
Thu (7–8pm IST)
Live mentor session · usually a working-pro problem swap + group code review
Sat (10–11am IST)
Live mentor session · weekly capstone check-in + senior code review
Sun
Optional · practice problems + your capstone work · solo, your pace
Fresher · 3–4 hrs/day · 5 days/week
Mon–Fri (10am–2pm IST)
Live cohort sessions at Hyderabad lab OR online · pair-programming · pre-recorded video review
Mon–Fri (afternoon)
Coding lab · mentor on the floor (offline) or DMing (online) · ~2 hours of focused build time
Fri (afternoon)
Demo day · ship what you built this week to your cohort + get senior feedback
Sat
Optional capstone day · open lab · mentors available for office hours
Sun
Off · rest, side projects, life
Every live session is recorded and dropped into your portal within 12 hours — so missing a session never means losing the content.
Your portal
Sign in. Enter activation code. Start learning.
After your Razorpay payment clears, we email you a one-time activation code. Sign in to the Nettms portal, paste the code, and the full course library unlocks instantly — 150 hours of pre-recorded modules, downloadable code, lesson PDFs, and your cohort’s live-session recordings.
- Pre-recorded video lessons · stream HD · 12-month access
- Downloadable code + module exercises + reference projects
- Live mentor session recordings (working pros) + daily class recordings (freshers)
- Capstone submission + verifiable Nettms certificate on completion
Activation-code gated
One code per enrolment · 12-month access · non-transferable
Common questions
Before you contact us, the answers everyone asks.
Ready to start the Applied AI Engineer Program?
₹50,000 · ₹35,000 with masterclass coupon · two tracks · 150 hours.
25-student cohorts · live + recorded · 12-month portal access