Put your own AI product online.
Build an application that uses documents and tools to complete a task. Test its answers and deploy a version people can try.
NYC Accelerator
AI Engineering
Build AI products you can crack internships with.
Build and deploy AI applications of your own. Get feedback from engineers and take a working product, test results and clear technical decisions into internship interviews.
Your instructors
Taught by engineers who got
the job you want.
They teach your live classes, review your projects and run your mock interviews.
How hiring works now
Interviews have changed.
Many companies now run two rounds: one without AI and one with it. NYC trains you for both.
No AI
Classic round
123456
- DSA and LLM basics, a timer, no AI
- Tests how you think on your own
Who fails here: people who need AI for everything.
AI allowed
AI-assisted round
1
answer = llm(question)2
if "refund" in answer:3
issue_refund(user)4
return answerAIAgent now handles refunds automatically.
YouA user can type "refund" and get one. No check.
- A codebase with an AI feature, AI open
- Tests whether you catch the unsafe step
Who fails here: people who only call an API.
At NYC, tests run without AI and projects run with it, so you practise both rounds.
What you’ll be able to show
Have an AI product to demonstrate.
Show the application, explain how it works and share the tests behind it. Build evidence of your engineering beyond a prompt.
Experiments you can explain
Show how you trained or evaluated a model and what you learnt from its failures.
An AI product you built
Connect models, documents and tools in an application someone can use.
A demo backed by tests
Show the test results and explain response time, cost and failure handling.
Duration8 months live + 6 months of recordings
Live classes5 a week, 1.5 hours each
Your toolsML · PyTorch · LLMs · RAG · Agents
Learn withMentors + teaching assistants
You build the work through assignments and projects. NYC gives you live classes, mentor feedback and teaching-assistant support.
Practise the decisions behind the project
How does an AI assistant
find the right answer?
Follow the decisions behind an AI answer. Learn to explain where it came from and how your application handles missing information.
Scripted concept demo · no live AI model
Can I return it after five days?
The item is unused. I have the receipt.
Retrieved evidence
Example policy / Returns: unused items may be returned within 7 days with a receipt.
Assistant response
First, retrieve the policy that answers this question.
1 questionOne clear task
A fluent answer is not enough. The assistant needs the store’s actual policy before it can help.
Interactive learning example. Illustrative data and outputs; not student performance or a live AI service.
Your portfolio starts here
Show what your AI can actually do.
Develop projects you can demo, explain and add to internship applications. These briefs show the kind of work you’ll practise.
Build an assistant that cites its sources.
Build an assistant that answers from documents and cites its sources. Show how you test incorrect answers and missing information.
Illustrative project brief
Document assistant
RAG + citations + evaluation set
Build agents with clear boundaries.
Build an assistant that uses tools to complete tasks. Demonstrate permissions, approvals and recovery when a step fails.
Illustrative project brief
Tool-using assistant
Tools + orchestration + failure recovery
Illustrative project briefs. Your course assignments and datasets are included in the batch plan.
How NYC helps you get there
Get your AI project reviewed by engineers.
Get help with the part you cannot solve. Get feedback on the decisions you made. Apply it to your next version.
Know how to start your own project.
Watch an instructor solve the problem and explain each decision. Then use the same approach in your own project.
Get past the error holding you up.
Bring your code and the error to a teaching assistant. Work out what went wrong and understand how to fix it yourself.
Get your project reviewed by an engineer.
Demo your project and explain your choices to a mentor. Find the gaps in your work and improve them before you put it in your portfolio.
A real NYC software project review
See what an engineer checks in a project.
Watch how NYC reviews a software project. In AI Engineering, reviews focus on your application, model behaviour, tests and technical decisions.
Learning scenes are illustrative. Your batch plan covers mentors, project reviews and support hours.
Beyond the first working demo
Show that your AI product holds up beyond the demo.
Test wrong answers, slow responses and failed tool calls. Give an interviewer evidence of how you checked and improved your product.
EvaluationDeploymentLatency + costMonitoring
Take an AI application into production.
Test the system’s failure cases, monitor latency and cost, and learn how to improve it after deployment.
Who teaches you
Working engineers, not full-time trainers.
They work in tech today, so you learn what teams use now.
Applied machine learning
Abhishek
Engineering Analyst, Google
IIT Kanpur. Earlier a data scientist at Optum and a scientist at BARC.
Agentic AI and AI engineering
Jaskirat
Senior Software Engineer, NTT DATA
Builds agentic AI platforms. Speaks on AI at Google and Microsoft events.
Instructors teach NYC live classes and review student projects. They speak for themselves, not for their employers.
What you’ll learn
The skills behind your AI portfolio.
35 weeks of live classes, from data foundations to deployed AI systems. Open a phase to see every week.
01 / Data foundationsWeeks 1–5 · SQL · Statistics · Python
01SQL and statisticsWeeks 1–3 · Queries · Window functions · A/B tests
SQL and statistics
- Week 1 · Data thinking and SQL I: querying and aggregating data
- Week 2 · SQL II: joins, CTEs and window functions
- Week 3 · Statistics: distributions, sampling and A/B tests
What you can do after thisQuery real data with SQL and reason about uncertainty.
02Python for dataWeeks 4–5 · pandas · Cleaning · Visualisation
Python for data
- Week 4 · Python for data: NumPy and pandas
- Week 5 · Cleaning messy data, exploratory analysis and visualisation
What you can do after thisUse Python to clean, explore and explain a dataset.
02 / Machine learningWeeks 6–16 · Maths · Models · ML engineering
03Framing and the mathsWeeks 6–7 · Framing · Linear algebra · Calculus · Inference
Framing and the maths
- Week 6 · Framing prediction problems, and linear algebra
- Week 7 · Calculus, optimisation and inference
What you can do after thisFrame a machine-learning problem and understand the maths models rely on.
04Machine learningWeeks 8–12 · Regression · Trees · Evaluation · NLP
Machine learning
- Week 8 · Linear and logistic models
- Week 9 · Trees and ensembles
- Week 10 · Features, data leakage and model evaluation
- Week 11 · Clustering, PCA and time series
- Week 12 · Text data, NLP fundamentals and a machine learning project
What you can do after thisBuild, evaluate and explain models on tabular data, time series and text.
05ML engineeringWeeks 13–16 · Pipelines · APIs · Docker · MLOps
ML engineering
- Week 13 · From notebooks to tested code; data and training pipelines
- Week 14 · Serving models behind APIs
- Week 15 · Docker, cloud and CI/CD for ML
- Week 16 · Monitoring, data drift and MLOps
What you can do after thisPut a trained model online as a reliable service and keep it healthy.
03 / AI engineeringWeeks 17–35 · Transformers · RAG · Agents · Production AI
06Deep learningWeeks 17–19 · Neural networks · Attention · Transformers
Deep learning
- Week 17 · Neural networks
- Week 18 · Representations and architectures
- Week 19 · Transformers and foundation models
What you can do after thisExplain what a transformer-based model is doing.
07LLM applications and RAGWeeks 20–22 · LLM APIs · Retrieval · RAG evaluation
LLM applications and RAG
- Week 20 · LLM application fundamentals
- Week 21 · Embeddings, vector search and RAG
- Week 22 · Advanced retrieval and RAG evaluation
What you can do after thisBuild retrieval-based AI systems and measure their quality.
08Adapting and evaluating modelsWeeks 23–24 · Fine-tuning · LoRA · Evals
Adapting and evaluating models
- Week 23 · Fine-tuning and adaptation (LoRA, PEFT)
- Week 24 · LLM evaluation engineering
What you can do after thisDecide when to fine-tune, and compare AI versions with evidence.
09Agents, multimodal and inferenceWeeks 25–27 · Tools · Vision · Speech · Serving
Agents, multimodal and inference
- Week 25 · Tool calling, workflows and agents
- Week 26 · Multimodal AI: images and speech
- Week 27 · Inference engineering and model serving
What you can do after thisBuild agents and multimodal features, and reason about inference cost.
10Safe, production AIWeeks 28–31 · Security · System design · Real users · Open source
Safe, production AI
- Week 28 · AI safety, security and reliability
- Week 29 · Production AI systems and system design
- Week 30 · AI product sprint with real users
- Week 31 · Open source or research-to-engineering sprint
What you can do after thisOperate AI products safely with real users.
11AI engineering final productWeeks 32–35 · Architecture · Build · Harden · Defend
AI engineering final product
- Week 32 · Architecture and evaluation plan
- Week 33 · Build and evaluate
- Week 34 · Production hardening
- Week 35 · Launch, explain and defend
What you can do after thisA live AI system, and proof you can change, debug and defend it.
Monday to Thursday: learn and practise. Friday: build, case, debug or review. Each live class is 1.5 hours.
Course fee
Your first internship starts with the work.
The course includes project work and learning support. Pay once, or in 12 monthly payments on EMI, with interest as low as 1% a month on the course fee. Add one month of internship preparation after the course, with a complimentary session with a MAANG engineer, for ₹15,000.
AI Engineering
Your selected plan
₹7,000/monthfor 12 months · interest as low as 1% a month on the course fee
Less than one semester of private engineering college fees.
- Course fee
- ₹75,000
- Internship preparation + MAANG engineer session
- ₹15,000
- EMI interest (1% a month on the course fee)
- ₹9,000
- Total
- ₹84,000
Included in the course
- Course lessons and guided practice
- Projects for your portfolio
- Project feedback from mentors
- Teaching-assistant doubt support
- NYC learning community
Optional support
See everything includedRésumé reviews, referrals, timed test practice and mock interviews with engineersEMI is 12 monthly payments, with interest as low as 1% a month on the course fee. The add-on has no interest.
What if I do the work and still don’t finish a live project?
You join the next batch’s live classes at no extra fee, once. Doing the work means attending at least 85% of live classes, watching every class you miss within 7 days, submitting every weekly task, attempting the practice work, and building and presenting your final project. A live project runs at a public link, its code is on GitHub, and you can explain it in your final review. Ask within 30 days of your course ending. Read the full rules.
Who is AI Engineering for?
College students who want to build AI products and prepare for AI engineering internships. Start with the foundations, then build applications with project practice and mentor feedback.
Is the add-on optional?
Yes. The ₹15,000 add-on is one month of internship preparation after your live classes end, plus a complimentary 45-minute session with a MAANG engineer. Tick it in the fees section to see your new total, or join the course without it.
Can I start from the foundations?
Prior AI experience is not required. The course starts with Python, data and mathematical foundations and builds towards advanced work. Build your understanding through coding, model evaluation and regular project practice.
Can I manage this alongside college?
Yes. Live classes are 1.5 hours, Monday to Friday, about 7.5 hours a week, plus project work. Fridays are build, case or review days. Missed a class? Watch the recording. You keep recordings for 6 months after live classes end.
What will I be able to show?
A portfolio of analysis, modelling and AI applications, with demos, test results and technical documentation. You build it through assignments, project reviews and practice. The briefs above are examples.
How will I get help?
Teaching assistants help with code and concepts. Mentors review your project decisions so you know what to improve. Your batch plan includes the support channels and review schedule.
How much does this course cost?
The course fee is ₹75,000, or ₹7,000 a month for 12 months on EMI (interest as low as 1% a month on the course fee). The optional ₹15,000 add-on is one month of internship preparation after the course, with a complimentary session with a MAANG engineer (₹8,250 a month on EMI with the add-on). An advisor shares the complete fee breakdown before you enrol.
Does this course guarantee an internship?
NYC helps you build practical skills and a portfolio you can explain. Dedicated internship preparation is available as an optional add-on. Employment is not guaranteed; employers make hiring decisions.
Join NYC
Take your own AI product into the interview.
Build the product and the understanding to explain it. Get the AI Engineering project plan, batch schedule and enrolment details.
On the call
01 / Your goals & starting point
02 / Curriculum, support & timetable
03 / Your plan, fees & joining