AI-150 · Program
Applied AI Integration
A practical program on using AI in products, not just in notebooks. You will train simple models, build an assistant on top of LLM APIs, and run a small model on a microcontroller.
6 weeksApprox. 90 hoursBeginner to intermediateClassroom + lab
Who this is for
- Students from any branch who know basic programming
- Developers who want to add AI features to apps
- Embedded engineers curious about edge AI
What you'll be able to do
- Prepare data and train basic ML models
- Call LLM APIs and get structured output
- Build a document Q&A assistant
- Run a small model on an ESP32 or STM32
- Judge cost, accuracy and privacy trade-offs
Syllabus
7 modules. Every module has lab sessions on real boards or tools.
M01Python for AI
- Python essentials
- NumPy, pandas and notebooks
- Plotting data
M02Machine learning basics
- Regression and classification
- Training, testing and metrics
- scikit-learn workflows
M03Working with data
- Cleaning and feature engineering
- Sensor and time-series data
M04LLM APIs and prompting
- Calling LLM APIs
- Prompt design and structured output
- Retrieval-augmented generation (RAG)
M05AI in products
- Adding AI to web and IoT apps
- Cost, latency, privacy and evaluation
M06Edge AI and TinyML
- TensorFlow Lite Micro and Edge Impulse
- Deploying models on microcontrollers
M07Capstone project
- AI feature in a working product
- Demo and review
Projects you can build
Predictive maintenance
Detect faults from vibration data.
Document Q&A assistant
Answer questions from your own PDFs.
Keyword spotting
Voice command recognition on a microcontroller.
Tools and technologies
- Python
- scikit-learn
- LLM APIs
- RAG
- TinyML
Every program includes
- Mock technical and HR interviews with working engineers
- Resume, LinkedIn and GitHub review
- Course completion certificate after the final assessment
- Referrals to our hiring network
Questions
Do I need strong maths?
Basic school maths is enough. We focus on using AI well rather than deriving algorithms.
Do I need a powerful laptop?
No. We use cloud notebooks and APIs, and small boards for edge AI.