Portfolio
Now Learning

My Skill Tree

The AI product engineer path — from a strong foundation up to the differentiators that make the role AI-proof. Go deep on Block B first; layer in the rest as projects demand.

45 owned45/62 · 73%

Mission

AI Product Engineer

A

Foundations

Owned · keep sharp

15/15
  • React.js
  • Next.js
  • TypeScript
  • JavaScript
  • Tailwind CSS
  • Redux / Redux Toolkit
  • Python
  • Django
  • Flask
  • REST APIs
  • Jest
  • React Testing Library
  • TDD
  • Git / GitHub
  • Razorpay / Payments
Priority
B

Core AI Specialization

Priority · this is the role

7/14
  • LLM APIs (OpenAI, Anthropic, Gemini)
  • Prompt engineering
  • Structured outputs
  • Function / tool calling
  • RAG (retrieval-augmented generation)
  • Embeddings & chunking
  • Vector DBs (pgvector, Pinecone, Qdrant, Weaviate)
  • Evals / LLM evaluation (offline + online, LLM-as-judge)
  • Guardrails (injection defense, content filtering)
  • AI agents & orchestration (multi-step, tool-using)
  • Agent frameworks (LangGraph, LlamaIndex)
  • Streaming / token-streaming UX
  • Cost & latency optimization (caching, routing)
  • Fine-tuning basics (know when not to)
C

Backend & Data Depth

Layer in as needed

5/7
  • FastAPI
  • Async / concurrency in Python
  • PostgreSQL (indexing, query optimization, transactions)
  • Redis (caching, queues)
  • API design
  • Background jobs / task queues (Celery)
  • Event-driven patterns (queues, pub/sub)
D

Reliability, Testing & Security

Layer in as needed

4/8
  • Structured logging
  • Tracing / OpenTelemetry
  • Monitoring & metrics
  • Retries
  • Idempotency
  • Fallbacks
  • Eval harness / regression tests for AI
  • Security basics (authz, secrets, OWASP, injection)
E

Cloud & Deployment

Layer in as needed

5/6
  • One cloud deep — AWS or GCP (pick one)
  • Serverless (Lambda / Cloud Functions)
  • Docker
  • CI/CD (GitHub Actions)
  • Deploy platforms (Vercel, Railway, Render, Fly)
  • IaC basics — Terraform / CDK (later)
F

Judgment & Product

Your AI-proof edge

4/7
  • Requirements → architecture
  • System design fundamentals (scaling, trade-offs)
  • Reading / verifying / debugging AI code
  • Code review
  • Product sense (what to build, AI feature UX)
  • One domain deep (e-commerce / fintech / proptech)
  • Technical writing
G

AI Coding Workflow

Daily tools

5/5
  • Claude Code
  • Cursor
  • GitHub Copilot
  • Codex
  • Agent delegation (task scoping, context mgmt)
Already own it Differentiator — go deep Priority block