Phase 1
Machine Learning Foundations
Weeks 1 to 3 establish the technical base through problem framing, Python for ML, baseline models, and evaluation.
The KalyrAITech Curriculum
A 10-week roadmap from machine learning foundations to RAG and introductory agents.
Learn machine learning, build LLM applications, and develop a document-based AI capstone across 20 live sessions.
- 10-week live cohort
- 20 guided sessions
- Midterm project and final capstone
Curriculum Overview
Four learning phases.
Phase 2
LLM Application Building
Weeks 4 and 5 cover practical LLM workflows, APIs, structured outputs, and reliable application patterns.
Phase 3
Retrieval and RAG Foundations
Weeks 6 to 8 focus on embeddings, semantic search, vector databases, document preparation, chunking, and grounded generation.
Phase 4
Evaluation, Agents, and Capstone
Weeks 9 and 10 improve quality, introduce bounded agent patterns, and complete the final capstone.
Week-by-Week Roadmap
What students learn and build each week.
Week 1
Machine-Learning Foundations and Python for ML
Problem framing, Python refresher, data handling, and the first ML repository setup.
Week 2
Building Machine-Learning Models
Supervised and unsupervised baselines, pipelines, and the first working ML notebook.
Week 3
Evaluation and ML Mini-Project
Metrics, error analysis, and an evaluated mini-project with a simple serving pattern.
Week 4
Large Language Model Foundations
Tokens, embeddings, prompting, model APIs, and structured output patterns.
Week 5
Building Reliable LLM Applications
Validation, safeguards, hallucination handling, and the midterm project launch.
Week 6
Embeddings and Semantic Search
Semantic similarity, embeddings-based search, and the first retrieval workflow.
Week 7
Vector Databases and Document Preparation
Vector storage, document extraction, cleaning, chunking, and indexed collections.
Week 8
Retrieval-Augmented Generation
Build a working RAG pipeline that answers from trusted source documents with grounded responses.
Week 9
RAG Evaluation and AI Agents
Improve retrieval quality, inspect failure cases, and add one bounded tool-calling pattern.
Week 10
Multi-Agent Systems and Capstone Showcase
Multi-agent fundamentals, final project demonstrations, architecture explanation, and interview preparation.
Milestone Progression
Five checkpoints along the way.
- End of Week 3: ML foundation with evaluated model workflow and basic serving.
- End of Week 5: Midterm project brief and launch, building on LLM application skills.
- End of Week 7: Retrieval foundation with semantic search, indexed documents, and repeatable ingestion.
- End of Week 9: Evaluated RAG prototype plus a controlled tool-calling extension.
- End of Week 10: Portfolio capstone with evaluation evidence and architecture explanation.
Next Step
Ready for the next cohort?
Review the entry requirements and application process before applying.