AI Software Engineer-For MTs

Start from zero and build the core programming foundation every AI software engineer needs. This module takes you from setting up your development environment to writing real programs in both Python and JavaScript — the two languages you'll use throughout the program. You'll learn to think computationally, work with data, collaborate using Git and GitHub, and understand how the web and APIs actually work. No prior coding experience is required; every concept is taught from the ground up with hands-on labs.

What you'll learn:

• Developer setup and computational thinking: installing Python 3.12+ and Node.js 22, VS Code, the terminal, problem decomposition and pseudocode
• Python core: variables, data types, strings and f-strings, conditionals, loops, lists, dictionaries and debugging
• Functions, modules, pip packages and virtual environments (venv); clean project structure
• Object-oriented programming, error handling, file I/O and working with JSON
• Data handling with NumPy and Pandas: loading, cleaning, transforming, grouping and plotting data
• JavaScript and TypeScript essentials: modern ES2024, the DOM, async/await, fetch and type basics
• Git, GitHub and collaborative workflows: branches, merges, pull requests and code review
• How the web works: HTTP, REST APIs, JSON, API keys and secrets

Turn your JavaScript foundation into modern, full-stack web applications. This module covers professional front-end and full-stack development with React and Next.js 16 — the same stack used to ship production apps today. You'll build responsive interfaces, manage application state, work with server components and databases, style polished UIs, and deploy your work live to the web. By the end you'll be able to build and ship a complete full-stack application.

What you'll learn:

• Web UI foundations: semantic HTML5, the CSS box model, Flexbox, Grid, responsive design and accessibility
• React core: JSX, function components, props, state with useState, events and controlled forms
• React hooks and state management: useEffect, useContext, useReducer, custom hooks, Zustand and TanStack Query
• Next.js 16 with the App Router: file-based routing, layouts, navigation and the metadata API
• Server Components, Server Actions, caching, streaming and Suspense
• Full-stack Next.js: API routes, PostgreSQL with Prisma/Drizzle ORM, CRUD and authentication with Auth.js
• Styling with Tailwind CSS and the shadcn/ui component library, dark mode and theming
• Deployment on Vercel: environments, Core Web Vitals, performance, SEO and analytics

Learn to build real applications powered by large language models. This module demystifies how LLMs work and teaches prompt engineering as a practical, repeatable skill. You'll work with both paid APIs (OpenAI, Anthropic, Gemini) and free/local models, build AI-powered apps with the Vercel AI SDK, and learn to secure and evaluate what you ship. The module closes with your first capstone prototype — a working LLM-powered application.

What you'll learn:

• How LLMs work: tokens, context windows, temperature/top-p, model families (GPT-5.x, Claude 4.x, Gemini 3.x, Llama 4) and hallucination
• Prompt engineering core: zero- and few-shot, role/persona, system prompts, delimiters and Chain-of-Thought
• Advanced prompting: self-consistency, ReAct, prompt chaining, JSON mode with Pydantic, tool calling and multimodal
• Building with LLM APIs: OpenAI, Anthropic and Gemini SDKs, streaming, cost management, plus local models with Ollama, Hugging Face and Groq
• Building AI apps with the Vercel AI SDK: streaming chatbots, conversation memory and generative UI
• Prompt security, evaluation and responsible AI: prompt injection, guardrails, PII, bias and quality evaluation
• Capstone project: an LLM prototype with prompted UX, system prompts and structured outputs

Give your AI applications access to real, up-to-date knowledge. This module covers Retrieval-Augmented Generation (RAG) — the technique behind "chat with your documents" and knowledge-grounded assistants. You'll learn how embeddings and vector databases power semantic search, then build a complete retrieval pipeline and tune it for quality. The module ends with a hands-on project connecting your capstone to a real knowledge base.

What you'll learn:

• Embeddings and semantic search: embedding models (text-embedding-3, Voyage, Cohere, BGE), cosine similarity, semantic vs keyword search
• Vector databases: HNSW indexing, quantization, metadata filtering; Chroma and pgvector (local), Pinecone and Qdrant (managed)
• RAG fundamentals: architecture, document loading, chunking strategies, indexing, retrieval and augmentation with LangChain/LlamaIndex
• Building a "chat with your docs" pipeline end to end
• Capstone project: build a document/data pipeline, create embeddings and vector indexes, and tune chunking, metadata filters and retrieval quality