How Surge Foundry works
A self-paced Pre-Flight, then a four-week Builder Sprint, then a Product Passport and a place in the Commons. Here's exactly what happens — followed all the way through by one real example.
At a glance
Apply + pick your chapter → Pre-Flight (get set up) → Week 1 Discover + Design → Week 2 Build → Week 3 Evaluate + Harden → Week 4 Verify + Ship → Product Passport → Demo Day → Surge Commons → Jobs, Founders or Research. Free, the whole way.
When you apply to a monthly cohort, you pick one chapter — your campus, your city, or a community like Women Builders, Developers, Career Switchers or Online India. Your chapter is your local crew; the cohort is the shared standard everyone is held to.
Meet Priya. She's a computer-science graduate in Chennai who knows some Python but has never shipped an AI product. We'll follow her through every step — and the product she builds: an AI agent that helps a small clinic route urgent patient messages to a human fast.
The journey
Five phases, four weeks
Get set up
Deliverable: Setup complete
Before the sprint starts, you get set up at your own pace — Git and GitHub, Python basics, Claude and coding-agent basics, APIs, environment setup, and core LLM ideas. That way Day 1 of Week 1 starts with building, not installing.
Priya spends a weekend on Pre-Flight, sets up her tools, and runs her first small agent.
Discover + Design
Deliverable: Product Blueprint
Orientation, then you choose a real problem, validate it, design the architecture, and design your AI agent. By the weekend you have a working prototype.
Priya scopes her clinic-triage agent and sketches how messages flow from intake to a human.
Build
Deliverable: Functional MVP
The main engineering week. You build your product with your own AI tools — backend, frontend, the AI agent, tools, RAG, database, APIs, basic testing — against daily build targets.
Priya builds her agent in github.com/priya-r/clinic-triage-agent. It's hers — during the sprint and long after.
Evaluate + Harden
Deliverable: Release Candidate
This is where Surge differs from a hackathon. You define what "works" means, benchmark it, test it, run basic security and responsible-AI checks, write the README, and keep your AI Engineering Journal.
When her agent mis-routes borderline cases, Priya works through three approaches with the AI, records the journey, and lands on one that lifts accuracy from 68% to 91%.
Verify + Ship
Deliverable: Product Passport
Peer review, then mentor verification against the five gates — with your individual contribution assessed. You present at Demo Day; pass verification and you earn your Product Passport, which can be issued in the weeks after Demo Day.
Priya's project passes all five gates; her individual contribution is verified. She earns her Product Passport and demos her agent.
Verification
How verification works
We don't give every project a personal mentor — that wouldn't scale, and it isn't needed. Instead your work passes through four layers: you self-check it, automated checks run, peers review it, and then a mentor from the pool verifies the projects that are ready. Mentors judge your product, engineering, AI, evaluation and contribution against a clear rubric — they don't read every line of code. It's fair, consistent, and it scales.
You confirm the five gates in the submission form.
The Pre-Flight Checker runs: README, licence, journal, benchmark, build, lint, tests, dependency scan.
Two peers review your project with a rubric. They catch obvious issues before it reaches a mentor.
Only projects that passed above reach a mentor. The weighted rubric decides — 80+ to verify.
After the sprint
Where it leads
Your verified product joins the Surge Commons — a permanent public library employers browse. From there, three doors open:
Job
Employers discover you by your evidence — your Product Passport, your Commons entry, your GitHub.
Startup
Validation, co-founders, mentors, and funding routes — for builders who want to start something.
Research
Publish, work with universities, contribute to the field — for builders who want to go deeper.
And you stay in the Surge Builder Network — many builders return as mentors.
Priya's agent appears in the Commons with a verified Passport. A recruiter finds it there — and next cohort, Priya comes back as a peer reviewer.
Before you apply
What you need
- To be 18 or older. No upper age limit.
- A GitHub account (free).
- Your own access to an AI assistant. We teach Claude-first, and the free tier is enough for the projects we scope — you do not need a paid plan to take part.
- Around 15-20 hours a week for the 4-week sprint, plus a few days of Pre-Flight.
- English, for now. The interface is English at launch.
What you leave with
Five things that compound
- A real, working AI product — in your own GitHub, yours forever.
- A verified Product Passport that proves what you built and what you contributed.
- An engineering journal that proves how you think.
- A place in the Commons where employers look — and three pathways out: jobs, founders, research.
- The one skill the market is hiring for: AI-native engineering.