Not just the roadmap — the architecture decisions, the team, the tradeoffs. Five weeks of hands-on practice building so you can shape decisions in every room. No coding required.
Illustrative — actual outputs depend on your product context.
20+ years building AI-native products at global scale — Meta, Microsoft, Yahoo!, Oracle. Ads, search, messaging. Works where business strategy meets ML infrastructure. 10 years teaching at ISB — from experience, not frameworks.
B.E., University of Mumbai · MBA, ISB
LinkedIn
4X founder. 20+ years building products across startups and big tech — Amazon, Yahoo, Sun Microsystems, Juniper Networks. Built two AI ventures — Mantid.ai and Juvant.ai. Has trained 600+ professionals.
CS, BITS Pilani · MBA, ISB
LinkedIn
3X founder. Built 91mobiles, SaaSworthy, and Intoglo across 20+ years in B2C, B2B, and cross-border logistics. Went deep on AI in 2021 — the models, the maths, and what separates real workflows from hype. Teaches with a builder's impatience for theory.
CEG, Anna University · MBA, ISB
LinkedInEngineering leads are running evals. Data science is choosing models. Platform teams are building the AI infrastructure. The PMs who understand these decisions are the ones shaping them — and the most valuable product leaders right now are those who can set AI strategy AND understand the systems well enough to execute it credibly.
This is about becoming the PM who sets your company's AI direction — who can write a credible AI product strategy, build and lead the team to execute it, and walk into any technical review with enough depth to shape the outcome. Strategy without execution depth is advice. This cohort gives you both. The technical depth is calibrated for what product leaders actually need: enough to engage meaningfully in architectural decisions and earn credibility in any technical room — no prior engineering background assumed.
Structured for working professionals — Wednesday evenings after the kickoff weekend. Each working session adds depth; self-paced Build Labs cement it on your own product between sessions, with weekly office hours to accelerate your build. Expect to invest ~40 hours across the programme.
The pedagogy here is deliberate: the only way to develop real AI product intuition is to build. Irrespective of your technical background, every session feeds directly into hands-on work — on your own product, your own use cases, your own roadmap. The bootcamp weekend gets you shipping from day one. By the capstone, you're demoing something live that you built yourself. No prior technical depth assumed. But by the end, you'll have it — because building is how you get there.
Oct 31 – Nov 1 · 14-hour weekend bootcamp. Hands-on Claude Code foundation that sets you up for everything that follows. Included in the full programme.
Nov 4, 11, 18, 25 · 7:30–10:30 PM IST. Three hours of live instruction, demos, and guided practice — culminating in a live capstone showcase on Dec 5.
2–3 hours of hands-on building each week. You ship something real — a working tool, a running agent, or a deployed prototype — applied to your own product, before the next session.
Weekly office hours with instructors. Bring your build, your questions, your blockers.
Dec 5 · 6–9 PM IST · The programme finale. Demo your live AI product or Agentic AI system to the full cohort — running code, not slides. Structured feedback from practitioners who've shipped real AI products.
Every learner ships a working AI product and demos it live, Q&A included. A few from the first two cohorts, unedited.
A co-pilot for first-time IndiaMART buyers: structured RFQs with AI gap checks, a chance-of-deal score for each seller conversation, prompt-injection flags, and an RFQ agent that picks the top sellers to negotiate with, plus model cost and eval comparisons.
A paper-trading portfolio builder on live Indian mutual fund NAV data: AI classifies the investor's risk profile and explains every recommendation, deterministic rules set the mix and guardrails, and backtesting and LLM-as-judge evals check the calls.
Detects when investors drift from their SIP goals, intervenes reactively from an intervention policy, and pre-empts drop-offs from alternate data such as account aggregator, credit bureau and app engagement signals, with an admin escalation queue and evals.
Two takes on one live problem from a GCC jewellery retailer: a monthly reorder engine that scores design velocity, respects manufacturing capacity and explains every line with AI, and persona-based flows for store, regional and new-collection managers, including a Bayesian borrowing model for designs with no sales history.
Graduates who complete the programme receive a credential that lives on a third-party blockchain — independently verifiable by anyone, for life, without needing to contact theCRUX. Share it on LinkedIn, attach it to a proposal, or link it from your profile. It can't be faked, revoked, or lost.
Cohort 4 starts November 4. Weekend bootcamp Oct 31 – Nov 1 · Wednesday sessions Nov 4 – Nov 25 · Demo Day Dec 5 · 25 seats per track.
All prices inclusive of taxes · Claude Pro ($20/month) + Anthropic API (~$5) required separately · GST invoice provided · 25 total seats across both tracks
All teaching and demonstrations run on Anthropic's Claude stack — Claude.ai, Claude Code, and the Anthropic API. Your own builds don't have to: for the build labs and capstone, work in whichever coding agent you already know — Claude Code, Codex, or Antigravity. If you build on the Claude stack, you'll need two things beyond the course fee:
Cohort 4 is open. 25 seats · early bird pricing. Full Programme ₹44,999 or Sessions Only ₹24,999.
Want to talk it through first? Email us at hello@thecrux.ai →