Questions builders ask.
Everything you need to know about Blink Labs, Phase 0, the 12-week program, costs, and outcomes.
Anyone with basic programming knowledge (any language) can apply. You don't need ML or AI experience — we teach that. You do need to be comfortable writing code and using a terminal. If you can write a simple Python script and use Git, you're ready for Phase 0.
No. We don't look at your CGPA, college ranking, or academic transcripts. Blink Labs selects for intent and commitment, not credentials. Your Phase 0 submission is what matters — did you show up, build something, and ship it on time?
Phase 0 is a free, 14-day builder challenge open to everyone who applies. You'll set up your development environment, learn terminal and Git basics, and build a simple CLI tool that calls external APIs. It's designed to test your commitment, not your credentials. There is absolutely no cost for Phase 0.
Plan for about 1-2 hours per day over 14 days. Some days will be lighter (setup, reading), some heavier (building and debugging). It's designed to be doable alongside college or a job, but you do need to be consistent. Missing multiple days usually means you won't finish in time.
The program is fully online with live sessions, mentor office hours, and async collaboration. All sessions are recorded if you miss one, but live attendance is strongly encouraged. We use Discord for daily communication and code reviews.
Expect 15-20 hours per week. This includes live sessions (3-4 hours/week), mentor office hours, and independent building time. The bulk of your time will be spent actually building your AI product. This is not a passive course — you're shipping every two weeks.
Every builder gets a dedicated mentor who checks in weekly. If you're struggling, we'll pair you with someone who can help. That said, the program is intentionally intense. If you consistently miss deadlines without communication, you may be asked to defer to the next cohort.
You pay nothing until after you complete Phase 0 and are accepted into the 12-week program. The program fee is ₹24,000, due only after Phase 0 approval. Merit-based scholarships are available — if cost is a barrier, reach out to us. We don't want finances to stop a committed builder.
Yes, builders who complete the program receive a Blink Labs completion certificate. But honestly, the certificate is the least valuable thing you'll walk away with. Your deployed AI product, your GitHub portfolio, and your Demo Day presentation are what actually get you hired or funded. Employers care about what you built, not a PDF.
The program is designed to make you job-ready as an AI engineer. You'll leave with a deployed product, a portfolio of real work, and experience with production tools (LangChain, Docker, Supabase, vector databases, etc.). Several alumni from past cohorts have landed AI engineering roles within weeks of Demo Day. We also connect top builders with hiring partners.
Absolutely — that's one of the highest outcomes we design for. The best Demo Day projects may receive support through our startup arm, including mentorship, introductions to investors, and help with early product-market fit. Multiple alumni projects have gone on to raise pre-seed funding.
Any laptop with at least 8GB RAM running macOS, Linux, or Windows (with WSL) will work. You'll need a stable internet connection for live sessions and deploying your projects. All compute-heavy AI work (model inference, training) happens on cloud infrastructure — you don't need a GPU.
Primarily Python and TypeScript/JavaScript. You'll work with LangChain, LangGraph, Docker, Supabase, vector databases (Qdrant, Pinecone), various LLM APIs (OpenAI, Anthropic, open-source models), and deployment tools like Vercel and Railway. We also cover observability tools like Sentry and Datadog.
Most AI courses teach you to follow tutorials. Blink Labs makes you build and deploy a real product. There are no pre-recorded lectures, no multiple-choice quizzes, no hand-holding. You get a problem, mentors who've shipped AI at Google/Meta/NVIDIA, and 12 weeks to build something real. The output isn't a certificate — it's a live, deployed AI system.
Still have questions? We'd love to hear from you.