✈
Welcome · Cohort 1

Twelve weeks to go from curious to capable, together.

About me

Hackathon winner4×
Currently helpingStartups
The roadmap

WK 1–3Python
WK 4–6Machine Learning
WK 7–8Deep Learning
WK 9–10LLMs & Agentic AI
WK 11–12Capstone
October 2026 · Weeks 1–4

Week 1 · October

Python for AI & AI Engineering

Python fundamentals
Functions, data structures, and modules
NumPy basics for arrays and vectorized computing
Notebooks & APIs · Jupyter/Colab basics + calling simple APIs
Virtual environments & packages · pip, requirements, reproducible setups
Git/GitHub workflow · clone, commit, push, pull requests
Intro AI/ML tooling · what we’ll use and how to set it up
Lab

Small Python data-processing app

Deliverable

Python/Colab project on GitHub

Week 2 · October

Data, APIs & Embeddings

NumPy & Pandas foundations · arrays, DataFrames, common operations
Data cleaning · missing values, duplicates, types, normalization
Working with real datasets · CSV, JSON, and basic dataset exploration
APIs/JSON · request/response, auth basics, parsing JSON
Text prep · tokenization basics, cleaning, chunking intuition
Embeddings intro · what embeddings represent and why they work
Vectors & similarity · cosine similarity, nearest neighbors intuition
Vector DBs · why they exist, how indexing/search works at a high level
Lab

Simple semantic-search system

Deliverable

Semantic-search notebook/app

Week 3 · October

ML → Modern AI

ML/DL basics · what models learn, parameters, and generalization
Supervised vs unsupervised learning · when to use each
Train/val/test splits · why we split and common pitfalls
Features · what features are and how they impact results
Neural nets · intuition: layers, activations, training loop concept
Transformers · high-level: attention idea and why it changed NLP
Pre-trained models · what “pre-trained” means and why we reuse them
Inference vs training · cost, compute, and when you do each
Evaluation · metrics, baseline comparisons, error analysis basics
Lab

Use a pre-trained model for a real task (stretch: compare models)

Deliverable

AI experiment notebook

Week 4 · October

Generative AI & AI Applications

LLM fundamentals · what an LLM is and what it can/can’t do
Tokens & context · context window, truncation, and prompt budgeting
Prompt engineering · instruction, examples, constraints, iteration
Structured outputs · JSON schemas, validation, and reliable formatting
Embeddings/semantic search · when to use embeddings vs generation
RAG fundamentals · retrieve → augment → generate workflow
AI agents & tool use · tool calling, multi-step reasoning loops
Evaluation · quality checks, regression tests, basic human eval rubrics
Responsible AI · safety, bias, privacy, and data handling expectations
Intro AWS AI/ML services · where managed services fit in the stack
Lab

Build a small AI app with an LLM/API

Deliverable

Working AI prototype + GitHub README

November – December 2026 · Weeks 5–11

November · Core builds
December · Capstones
Week 5 · November

Production RAG

Chunking strategies · size, overlap, semantic chunking tradeoffs
Retrieval quality · recall vs precision, what “good retrieval” means
Reranking · why rerankers help and where to apply them
Vector DB tuning · indexes, filters/metadata, and query patterns
Evaluating retrieval · test sets, retrieval metrics, and qualitative checks
Lab

Rebuild Week 2's semantic search into a real RAG pipeline over a larger dataset

Deliverable

GitHub milestone 1: RAG pipeline

Week 6 · November

AI Agents & Tool Use

Function calling · tool schemas, inputs/outputs, and guardrails
Multi-step agents · planning, executing, and stopping conditions
Agent frameworks · LangChain/LlamaIndex or AWS Bedrock Agents — what they provide
Tool orchestration · routing, retries, and chaining multiple tools
Memory · short-term vs long-term memory, and when it helps
Lab

Build an agent that chains 2+ tools to complete a task

Deliverable

GitHub milestone 2: working agent

Week 7 · November

Shipping AI Apps

Wrapping models/agents as APIs · FastAPI/Flask patterns and endpoints
Containerization basics · Docker images, environments, reproducible deploys
Calling from a simple frontend · client/server basics and request flow
Lab

Turn an existing lab into a callable API

Deliverable

GitHub milestone 3: deployed API

Week 8 · November

AWS for AI Engineers

Deploying to AWS · Lambda, SageMaker endpoints, Bedrock — when to use what
Cost/latency tradeoffs · batch vs real-time, caching, scaling
Monitoring basics · logs, metrics, tracing, and alerting signals
Lab

Deploy Week 7's API on AWS

Deliverable

GitHub milestone 4: AWS-hosted app

Week 9 · December

Capstone scoping & build

Students pick a real AI application problem, start building

Deliverable

Capstone scoped + in progress

Week 10 · December

Build sprint

RAG/agent/GenAI integration, iterate

Deliverable

Feature-complete build

Week 11 · December

Deployment & evaluation

Deployment, evaluation, responsible-AI review, README/documentation

Deliverable

Deployed capstone

What you'll gain

Technical

  • Python & data
  • ML & deep learning
  • LLMs & agents

Engineering

  • Git & collaboration
  • Cloud on AWS
  • Shipping & testing

Professional

  • Communication
  • Teamwork
  • Presenting your work
Rules every student must follow

  • 2 absences without permission is bye bye.
  • Inform the team in advance if you cannot attend.
  • Repeated unexcused absences trigger a warning and may lead to removal from the cohort.
  • Submit assignments by the deadline.
  • If you need an extension, request it before the deadline with a valid reason.
  • Repeatedly missing deadlines without communication affects your standing in the cohort.
  • Complete hands-on exercises, not just tutorials.
  • Push your work to GitHub regularly.
  • Each student must demonstrate individual understanding, even when working in a team.
  • Use AI to learn, debug, and explore ideas.
  • Never submit work you cannot explain.
  • Be honest about your contribution and give credit to others.
  • Ask questions, share progress, and help your peers.
  • Don't disappear when work gets difficult.
  • If you're stuck, explain what you've tried and ask for help.
  • No harassment, disrespect, plagiarism, or taking credit for others' work.
  • Keep discussions relevant and constructive.
  • Respect mentors, teammates, and different skill levels.
Consequences

Stage 1Warning

Private conversation, identify the issue, and agree on a corrective action.

Stage 2Probation

A clear improvement plan with specific tasks and a deadline. Review attendance, submissions, and participation.

Stage 3Removal from the cohort

For continued failure to meet requirements, serious misconduct, or academic dishonesty, following a fair review process.

Opportunities

MentorshipProjectsHackathonsWorkshopsShowcasesNetworkingLeadershipCertification ★
Community

You get

  • Mentors & peers
  • Real projects
  • AWS resources

We expect

  • Show up weekly
  • Build and share
  • Help each other

How to join in

  • Attend sessions
  • Join a squad
  • Demo at showcases
Mentors

Zahir Ahmed Seid
ZATap to add photo
Zahir Ahmed SeidMentor
Robel Ermiyas
RETap to add photo
Robel ErmiyasMentor
Lidiya Abebe
LATap to add photo
Lidiya AbebeMentor
Your goals

What do you want to achieve this cohort?

Thank you

Join this if you want to get to know me.

Scan · Telegram channelt.me/nrdNet

Control from your phone

Scan with your phone camera. Keep this tab open while you present.

QR code

Waiting for your phone…

RemoteConnecting…

Fully paid · Active students

Get AWS Certified.

Active students receive a fully paid AWS Skill Builder certification program, with a real chance to become an AWS Certified Cloud Practitioner.

  1. Stay active in the cohort
  2. Prepare through AWS Skill Builder
  3. Take the exam and earn your badge
AWS mascot
Scroll ↓