150-Day AI Challenge // Zero to AI Master

NULL
VECTOR

nullvector.monster

You know nothing about AI right now.
In 150 days you will build real AI systems, earn industry certifications, and implement a transformer from scratch.
Every day has a mission. Master it before it masters you.

150
Days
7
Phases
14
Harvard Projects
5+
Certifications
0
Prior Experience
Download Phase 1 → See All 7 Phases
150
Python · PyTorch · NumPy
Harvard CS50 AI · 14 Projects
Transformers · Attention · LLMs
Fine-Tuning · LoRA · RLHF · RAG
Agentic AI · MCP · LangChain
AWS ML · Azure AI · NVIDIA NCA
MLOps · Docker · CI/CD
150 Days · Zero to AI Master
Python · PyTorch · NumPy
Harvard CS50 AI · 14 Projects
Transformers · Attention · LLMs
Fine-Tuning · LoRA · RLHF · RAG
Agentic AI · MCP · LangChain
AWS ML · Azure AI · NVIDIA NCA
MLOps · Docker · CI/CD
150 Days · Zero to AI Master
Why This Challenge Exists

THE CHALLENGE
BEHIND ALL OF IT

The more people who understand AI, the more problems get solved. The more problems that get solved, the better people's lives become. It is that simple and that serious. NULL VECTOR exists to put this technology in the hands of as many people as possible — not as consumers, but as builders.

AI built by people who care about people is the difference between technology that extracts value and technology that creates it. Every person who completes this challenge is one more person capable of making something better — for a community, a family, a stranger, a problem that has been waiting too long for someone with the right tools to arrive.

The 150-Day Journey

7 PHASES.
ZERO SHORTCUTS.

Each phase is 21 days. Each phase ends with a verifiable milestone — a certification number or a deployed URL. Miss it and you restart that phase. No extensions. No partial credit.

01
Days 1–21
IGNITION — Harvard CS50 AI + Python
Complete all 14 Harvard CS50 AI projects — BFS, Minimax, Bayesian Networks, CNNs, BERT — while learning Python from scratch. Two certifications by Day 21.
⬡ Harvard Cert
CS50 AI Certificate +
Azure AI-900
02
Days 22–42
SIGNAL — Classical Machine Learning
Scikit-Learn, PyTorch, gradient descent from scratch, 5 ML models. No black boxes — every parameter has a reason. Phase ends with a live deployed web application.
⬡ Live App
Deployed ML Classifier
on HuggingFace Spaces
03
Days 43–63
DEPTH — Deep Learning
CNNs, RNNs, backpropagation derived by hand. fast.ai parallel track. GPU training. By Day 63 you can explain why model architecture decisions matter — and defend yours.
⬡ NVIDIA Cert
NVIDIA Certified
Associate: AI Infrastructure
04
Days 64–84
TRANSFORMER — LLM Architecture
Read Attention Is All You Need. Implement multi-head self-attention from scratch. Build and deploy a mini-GPT trained on real text. The hardest phase. The most important one.
⬡ Live LLM
Trained Language Model
Deployed Publicly
05
Days 85–105
OPERATOR — Fine-Tuning & RAG
LoRA, QLoRA, DPO, RLHF fundamentals. RAG pipelines with LlamaIndex. Fine-tune an open-source LLM on custom data. Two production deployments. AWS ML Specialty.
⬡ AWS Cert
AWS ML Specialty
MLS-C01
06
Days 106–126
AGENT — Autonomous AI Systems
LangChain, MCP protocol, multi-agent orchestration, n8n automation. Build an autonomous AI agent that completes multi-step tasks without human intervention. Deploy it.
⬡ Live Agent
Autonomous Agent
with Public URL
07
Days 127–150
DEPLOYMENT — MLOps, Portfolio & Job Attack
MLOps, Docker, CI/CD, cloud deployment. Package 7 projects into a portfolio. Write case studies. Record demos. Day 150: send 10 applications to AI companies.
⬡ Final Cert
Azure AI-102 +
10 Applications Sent
The Transformation

DAY 1 VS DAY 150

Day 1 — Where you start
—Never written a line of code
—AI is something that happens to you
—No certifications
—Cannot explain what a neural network does
—No public evidence of your skills
—No AI company would look at your resume
Day 150 — Where you arrive
✓150 consecutive GitHub commits
✓Harvard CS50 AI certified
✓Azure AI-900, AI-102, AWS ML Specialty, NVIDIA NCA
✓Built and deployed a transformer from scratch
✓Autonomous AI agent with public URL
✓Portfolio of 7 live AI applications
✓AI is something you build
How It Works

YOUR DAILY
PLAYBOOK

01
Write code every day
Even just a few lines. Push it to GitHub. The habit of showing up daily is more important than how much you do on any single day. Small steps compound into big skills.
02
Try it before you study it
After watching a lecture, close the video and attempt to build what you just saw. Getting it wrong is fine. Trying first is how the concept actually sticks.
03
Each phase has a finish line
Every 21 days there is a concrete milestone — a certification or a live deployed project. Finish lines keep you moving forward and give you something real to show for your effort.
04
Ship something people can use
The goal of each phase is a working application anyone can access via a URL — not just code sitting on your laptop. Shipping is what makes it real, and real is what gets you hired.
05
Write down what you learned
Three sentences a day in a public GitHub journal. What you built, what confused you, what you'll tackle tomorrow. Writing it out cements the learning and builds a record of your growth.
06
Use AI to learn, not to skip
Claude or ChatGPT are incredible tutors — ask them to explain errors, clarify concepts, and suggest approaches. Just don't ask them to do the work for you. The skills need to live in your hands.

150 DAYS.
AI MASTER.

Download your phase curriculum. Every day mapped. Every resource linked. Every milestone defined. No email required.

01
IGNITION
Days 1–21 · Harvard CS50 AI · Azure AI-900 · 21 Days Mapped
↓ PDF
02
SIGNAL
Days 22–42 · Classical ML · Live App Deployed
↓ PDF
03
DEPTH
Days 43–63 · Deep Learning · NVIDIA NCA-AIIO
↓ PDF
04
TRANSFORMER
Days 64–84 · Build an LLM from Scratch · Deploy It
↓ PDF
05
OPERATOR
Days 85–105 · Fine-Tuning · RAG · AWS ML Specialty
↓ PDF
06
AGENT
Days 106–126 · Autonomous AI · Live Agent App
↓ PDF
07
DEPLOYMENT
Days 127–150 · MLOps · Azure AI-102 · 10 Jobs Applied
↓ PDF
// All 7 PDFs free · No sign-up required · Start Day 1 today