Build production
AI agents.
Get hired.
The 8 weeks MasterClass that trains you the way a startup would — ship a real multi-agent system, not a tutorial project.
Next cohort: June 2, 2026
50 days away — runs quarterly
Shopify
Stripe
AnalyticsThis is not another AI course
Most programs teach you to call an API and build a demo. We train you to ship a production system the way a startup would — under real constraints, with real tools, on a real deadline.
“Imagine you just joined a startup as the first AI engineer. You have 8 weeks to ship. That’s this program.”
| Typical Course | This MasterClass | |
|---|---|---|
| Curriculum freshness | Recorded once | Rebuilt each cohort |
| Project scope | Single-file scripts | Full-stack system |
| Deployment | Localhost demo | Cloud + CI/CD |
| Evals & testing | Skipped | 50+ eval cases |
| Team experience | Solo work | Startup simulation |
| Security | Not covered | PII, RBAC, OWASP |
2026 best practices
The AI stack changes every quarter. Our curriculum is rebuilt each cohort around the tools and patterns that are actually winning in production right now — not what worked 12 months ago.
Real systems, not toy problems
You won't build a chatbot that answers FAQ. You'll build a multi-agent Financial Intelligence system with real data sources, evals, guardrails, and cloud deployment.
Full-stack, end-to-end
From retrieval engineering to FastAPI serving to Kubernetes deployment. You own the entire stack — the same scope you'd face as the first AI engineer at a startup.
Startup simulation, not lectures
Phase 4 is a team Startup Project. You pitch, build, and present a production system on Agent Demo Day — the closest thing to joining an early-stage AI company without the equity risk.
What you'll build in 8 weeks
You won't just learn concepts — you'll ship real systems. Every lab produces a portfolio-ready project you can show employers.
SQL Agent
Natural-language to SQL. Ask a question in English, agent generates the query, executes it, and returns a formatted answer.
| Product | Revenue | Region |
|---|---|---|
| Pro Wireless | $284K | US |
| Ultra Case | $198K | EU |
| Smart Dock | $156K | US |
| PowerPack | $134K | APAC |
| AirMount | $121K | US |
RAG Pipeline with Re-ranking
Hybrid search (vector + BM25) with cross-encoder re-ranking. Context engineering that actually retrieves the right documents.
Custom MCP Server
Build a Model Context Protocol server that connects LLMs to your local filesystem, databases, and internal APIs.
Image-to-Structured Data Agent
Multimodal agent that takes an image (receipt, invoice, screenshot) and extracts structured data with Pydantic validation.
Financial Intelligence Agent
The flagship project. Multi-agent system with Shopify, Stripe, and analytics integrations. Deployed to cloud with evals and CI/CD.
Startup Project
Your team picks a domain, pitches the idea, builds it end-to-end, and presents on Agent Demo Day to the cohort and guest reviewers.
The Financial Intelligence Agent for eCommerce
By graduation, you'll have built and deployed a system like this — a multi-agent pipeline that connects to live data sources, reasons across them, and delivers answers with traceable citations and eval scores.
Spring Collection
Up to 40% off select items
Free shipping on orders over $75
Sneakers Edition
Classic T-Shirt
Portable Speaker X
Only 3 left
Conversion rate down 18%
US storefront · Last 7 days
2.6%
was 3.8%
The 8-Week Curriculum
4 phases. Live sessions (Tuesday & Thursday) + offline labs. Culminates in a team Startup Project and Agent Demo Day.
Agent Fundamentals
Goal: Master the building blocks — retrieval, tool use, and multimodal AI.
Architecture & Context
AI-native development, 2026 AI Stacks, LLM Provider, ReACT.
Retrieval, chunking, re-ranking.
Build an agent that takes a natural-language question, generates SQL, runs it, and returns a formatted answer.
Tools & Multimodal
Type-safe tools, error handling, building a financial tool registry. MCP deep dive: build a custom MCP server.
Lab — Image-to-Structured Data Agent.
Build a Logo Generator Agent.
Data Analyst Agent
Goal: Compose fundamentals into a real product — a Financial Intelligence Agent.
Multi-Agent Design & Memory
Financial tool registry (market data, SEC EDGAR, news, internal KB). Structured output enforcement. Extended thinking vs. tool calls.
Thread persistence (Postgres/Redis). Context summarization. Personalization: the system learns analyst preferences, coverage universe, format.
Expand eval suite to 50 cases. Implement LangSmith/Langfuse tracing. Build cost dashboard ($/request by agent step).
Evals & Deployment
Write 15 eval cases: 5 factual accuracy, 5 citation quality, 5 hallucination detection. Run them. Find failure modes you didn't know about.
Docker for multi-service agents. FastAPI serving (SSE, async). Deploy to Railway/Vercel/Supabase. CI/CD with GitHub Actions: evals on every PR. Health checks, logging, Sentry.
Deploy the FIA Agent to production infrastructure.
AI Ops
Goal: Optimize cost, performance, and security for production systems.
Performance & Open-Source Models
Prompt caching (80-90% cost cut). Model cascading. Semantic caching. Batching & coalescing. Token optimization & prompt compression. Latency budgeting.
Build vs. buy framework. Llama, Mistral, Qwen, Gemma. PEFT/LoRA fine-tuning. Dataset curation from production logs. vLLM deployment. Quantization (GGUF/AWQ). Self-hosting breakeven math.
Apply cost optimization and model cascading to the Financial Intelligence Agent.
Security & Compliance
PII redaction middleware. Prompt injection defense. Audit trails with provenance. RBAC for data access. Zero-retention API patterns. OWASP AI Top 10 walkthrough.
Startup Project
Goal: Apply everything to a team project and present on Agent Demo Day.
Pitch Day
Teams pitch their project idea, architecture, and plan to the cohort.
Sprint Week
Teams build their agentic system end-to-end.
Progress review, architecture feedback, blocker resolution.
Demo Day
Demo DayPolish, test, prepare presentation.
Each team presents to the full cohort + guest reviewers. Live demo, architecture walkthrough, eval results, cost analysis, Q&A.
Each 60-Minute Live Session
Concept Deep Dive
The "Why"
Live Build-With-Me
Coding in a shared repo
Breakout Challenge
Small groups solve a task
Architecture Review & Q&A
Review and open discussion
Skills that get you hired
Every skill maps directly to what employers are hiring for right now. This isn't theory — it's the stack behind the job postings at top AI companies.
Context Engineering
RAG, Chunking, Re-ranking, Vector Search
Agent Systems
LangGraph, Tool Calling, MCP, Multi-Agent
Software Eng
Python, FastAPI, Pydantic, Async
Evals & Testing
DeepEval, RAGAS, LLM-as-Judge
Infra & Deploy
Docker, K8s, CI/CD, Cloud
AI Ops
Cost Optimization, Caching, Model Cascading
Security
PII Redaction, RBAC, OWASP AI Top 10
Model Tuning
LoRA/PEFT, vLLM, Quantization
Tools & frameworks you'll work with
LLM Providers
Commercial & open-source models
OpenAI
Claude
Deepseek
Gemini
Kimi
LlamaAgent Frameworks
Orchestration & serving
Python
LangChain
FastAPI
Hugging Face
OllamaData & Retrieval
Vector stores, caching & search
Pinecone
Redis
PgVector
Qdrant
Weaviate
MongoDB
Elasticsearch
SQL
MilvusInfra & Ops
Deploy, monitor & iterate
Opik
Docker
Kubernetes
GCP
GitHub
VS CodeJoinAI
Certificate of Completion
Startup AI Engineer
MasterClass
Awarded to
John Smith
Summer 2026 Cohort · Issued August 2026
Dan Lee
Instructor & Tech Lead
ID: JA-2026-0847
joinai.com/verify
Prove you can ship
This isn't a completion badge for watching videos. You earn this by building and presenting a production-grade AI system on Agent Demo Day — the same proof employers actually care about.
Verified completion
Issued upon completing all 4 phases, labs, and Agent Demo Day presentation.
Portfolio-ready
Comes with a public profile page listing your projects, skills, and capstone.
Shareable credential
Add to LinkedIn, resume, and GitHub. Includes a unique verification link.
Graduate with a portfolio, not just a certificate
Every lab and project becomes a public repo on your GitHub. By Week 8, you have 5+ production-quality projects that hiring managers can actually review — not tutorial code, real systems.
Data Scientist → AI Engineer | JoinAI MasterClass Graduate
Pinned
Natural-language to SQL agent with tool calling, error recovery, and formatted output.
Multimodal agent that generates brand logos from text descriptions using image generation APIs.
Multi-agent system with RAG, SEC EDGAR integration, market data tools, evals, and cost dashboard.
Prompt caching, model cascading, semantic caching, and cost optimization patterns for production AI.
Team capstone: full-stack agentic system pitched, built, and presented on Agent Demo Day.
56 contributions in the last 8 weeks
Daniel Lee
AI Engineer & Consultant·Formerly at
For the past 10 years, I've worked as an AI, ML, and data science practitioner — previously at Google. Now I run AI projects as an external consultant for startups, developing and deploying scalable AI solutions.
As your instructor, I'll guide you through the entire stack — from retrieval engineering to multi-agent orchestration to production deployment. Every session is live-coded, not slides.
What graduates are saying
Real feedback from real engineers.
“I highly recommend Dan's course. He simplifies complex concepts, making the material accessible and engaging. What stood out was Dan's commitment to supporting each person's learning journey.”
Sara A
Data Science Manager
“Truly transformative. I built and deployed a financial agent using LangChain, Redis, and GCP with real-time tools like Google Search and SQL queries, with tracing and evaluation via LangSmith.”
Chaitanya K
Graduate at Wharton
“Over five weeks, I gained practical experience designing and deploying AI agents. I now feel confident applying these skills in real projects. Dan's teaching is approachable and grounded.”
Priyanka A
Senior ML Engineer
“A game-changer. Clear, practical, real-world teaching made complex concepts easy to apply. With his hands-on guidance I built and deployed AI systems and gained the confidence to use these skills immediately.”
Anushree J
Senior Data Scientist
“Dan demonstrates how popular libraries and frameworks work under the hood. Knowledge like this makes you confident in developing any AI systems because you understand the actual working principles.”
Vincent R
Software Engineer
“I not only built an agent during the course but also created and deployed my own as a capstone project. It was an empowering experience that gave me the skills and confidence to build AI solutions independently.”
Siddhi J
Software Engineer
“Dan's AIE Masterclass transformed me from an AI enthusiast into a true builder and practitioner. I learned to architect production-grade systems that go far beyond prototypes. Using LangGraph and GCP, I built a platform that cut manual analysis time from 40 hours to just 5 minutes.”
Bashir Sadat
AI Engineer
“I really enjoyed the course content, especially the hands-on approach that walked us through the full end-to-end process of AI engineering. Learning to use LangSmith, exploring model security, and understanding evaluation techniques were highlights for me. Overall, this course has pushed me to grow and think more critically as a Generative AI Engineer.”
Kiril Simov
Generative AI Engineer
Sara A.
Data Science Manager
Chaitanya K.
Wharton Graduate
Priyanka A.
Senior ML Engineer
Anushree J.
Senior Data Scientist
Vincent R.
Software Engineer
Siddhi J.
Software Engineer
Bashir Sadat
AI Engineer
Dan Lee
Tech Lead
Kiril Simov
Software Engineer
Join the Talent Collective
Enrollment isn't just a course — it's a permanent membership in a vetted community of AI engineers. Alumni connect with hiring managers, refer each other into roles, and collaborate on projects long after graduation.
Join the Startup AI Engineer MasterClass
Our next cohort starts June 2, 2026. Invest in skills that open doors to AI engineering roles averaging $160K–$200K+.
Builder
The core MasterClass. 8 weeks of live sessions, labs, and a team Startup Project — everything you need to build and deploy production AI systems.
$2,497USD
What's included
- All live sessions + offline labs
- 4 phases, 8 weeks of content
- Direct instructor Q&A every session
- Cohort of max 20 students
- Breakout challenges in small groups
- Team Startup Project + Agent Demo Day
- Certification of completion
- Lifetime access to recordings
- Discord community + alumni network
Coaching
Everything in Builder, plus 5 hours of private 1:1 sessions with the Tech Lead. Tailored to your domain, your projects, and your career goals.
$3,497USD
Limited to 5 students per cohort
What's included
- Everything in Builder
- 5 x 1-hour 1:1 sessions with Tech Lead
- Personalized project scoping
- Resume & portfolio review
- Career strategy session
- Direct Slack access to instructor
- Priority Q&A in live sessions
Career
The complete package — MasterClass + 1:1 coaching + full interview prep. Build the skills, then land the $160K+ role.
$4,497USD
What's included
- Everything in Coaching
- ML/AI Interview Class enrollment
- System design interview prep
- ML coding interview practice
- Behavioral interview coaching
- Mock interviews with feedback
- Job referral network access