Senior Product Leader · 250M+ Device Scale · AI Product Builder · 0-to-1
I used to need a team and a quarter to ship a product. Now I need a weekend and an agent.
-orange?style=for-the-badge){kind=icon}
ISB MBA (Flagbearer Award, top 1%) and most recently Group PM at InMobi (Jan 2023 – Apr 2026). I led Go-To-Market and 0-to-1 product across 8 markets on Glance AI Lock Screen — an AI-powered platform pre-installed on 250M+ devices via OEM and telco partnerships. I ship AI products on my own time at github.com/aksheyw, with focused AI and agentic certifications from Vanderbilt, Duke, and Outskill.
Most senior PMs have one or two of these — I've shipped all four:
- 📱 Consumer AI scale — Group PM on Glance AI Lock Screen (InMobi, Jan 2023 – Apr 2026), 250M+ devices across 8 markets. Led US market entry 100K → 10M devices (100X in 18 months), reaching 6M DAU; ARPU $10 → $15 (+25% YoY); D60 retention 49% → 55%.
- 💸 Fintech P&L — built Magicbricks Pay Rent from 0 → INR 50Cr GMV (100% MoM at peak), payment success 95% → 99%, settlement 48hrs → 2hrs, integrated 3 payment gateways with cashback wiring across 5 banks (HDFC, ICICI, Kotak, Standard Chartered, IDFC).
- 🤝 Fortune 500 partnerships — OEM: Samsung, Motorola, Sharp, Xiaomi, Realme. Telco: Verizon (US), SoftBank (Japan).
- 🛠 Hands-on AI builder — ship real products with Claude Code, Codex, n8n, Vapi, and compatible LLMs from OpenAI, Anthropic, and Google (Gemini). The repos below are my actual config, sanitized and shared.
Short written breakdowns of how I think about the products I build — the problem, the decisions and trade-offs, and what I'd do next:
- JobMagnet — making a portfolio feel like it belongs to the company you're applying to; brand as a rendering problem, not a data problem; five agents over one prompt.
- GrowFlowAI — removing the capture step entirely; meeting users where they already work; designing engagement with substance.
- Kahaani AI — moving the constraint from "can we afford to try this" to "what should we try"; honest cost transparency over impressive numbers.
- Ship Gate — putting a quality gate where the skip flag doesn't exist; knowing when to stop hardening; learning from a real footgun.
- Mission Mode — a self-directed PM exercise on growth strategy for English learners in India: segmentation, JTBD, growth model, working prototype.
| Mission Mode case study · live → | Self-directed PM exercise on growth strategy for English learners in India. Segmentation, JTBD, wireframes, growth model, working prototype. The artifact recruiters usually ask for, made public. |
| JobMagnet · gallery → | Paste a job description + your résumé → a 5-agent pipeline (research → brand → narrative → pitch → code) builds a portfolio site brand-themed for that company, then hands you a deployable Next.js project to download. Built in 7 days for the OpenAI × Outskill hackathon. Codex SDK + Next.js + Supabase + Hostinger VPS · ~96s end-to-end. |
| GrowFlowAI · demo → | AI task-management. Meeting notes become tasks via Telegram, web, or email. React + Supabase Edge + Capacitor (Android) + n8n + OpenAI (transcription, vision, LLM extraction). CI green, used daily. |
| Kahaani AI · demo → | 3 publication-ready audio scripts in 90 seconds — for ₹2 vs ₹2,000. Built in 4 days. React + Vercel serverless + OpenAI. |
| claude-code-ship-gate | A pre-push quality gate for Claude Code. It runs your tests, code review, security, and a secret scan before any push to your protected branch and blocks it until they pass, so a git push --no-verify can't slip past. Default-on for a personal install, opt-in from the marketplace; say "ship it" to run it. Hardened by a 14-lens review + 3 red-team passes. Install: /plugin marketplace add aksheyw/claude-code-ship-gate. |
| claude-code-deep-review | 14-lens iterative review methodology. Found 14 production bugs (2 ship-stoppers) on first use. |
| claude-code-pm-agents | 7 product-builder subagents — PM, growth, brand, ASO, SEO, YouTube, comms triage. |
| claude-code-rules | Opinionated global rules: honesty / earned-confidence, TDD, immutability, branching. |
| claude-code-learned-skills | 3 Docker / SSH / VPS skills auto-extracted from real debugging sessions. |
| context-bridge | Per-project wiki + auto-generated handoff prompts that stop cross-session amnesia. Install: npx skills add aksheyw/context-bridge. 5 commands, 11-step save+sync, honesty rules embedded in adopters' CLAUDE.md. v0.1.2 shipped with matrix CI green on Ubuntu + macOS. |
| career-command-center-template | Plugin template for an AI-native job-search workflow — 10 skills, marketplace install (/plugin marketplace add aksheyw/career-command-center-template), example resume PDF in examples/. |
I treat Claude Code like a real product surface, not a chat box. Every project I ship has its own subagents, rules, hooks, and skills — built and refined across multiple launches.
The honesty rule alone has changed how I work with Claude Code more than any prompt-engineering trick — it forces Claude to default to "I haven't verified that — want me to check?" instead of confident-sounding fabrication.
AI & Agents
Build
- InMobi Real Star Award (2× winner: 2022, 2025) — company's highest individual contribution award
- InMobi top-performer recognition — 4 consecutive years (2022-2025)
- ISB Flagbearer Award (October 2021) — Top 1% of MBA class
- ISB Alumni Association Board Member (2024-Present)
Focused AI and agentic certifications from Vanderbilt, Duke, and Outskill:
- Claude Code: SE with GenAI Agents — Vanderbilt University (2026)
- AI Product Management — Duke University (2026)
- Agentic AI and AI Agents for Leaders — Vanderbilt University (2026)
- Generative AI Strategic Leader — Vanderbilt University (2025)
- AI Generalist Accelerator Program — Outskill (2025)
- Digital Product Management Specialization — UVA Darden
Exploring senior product roles · AI Product Building · 0-to-1 GTM at AI-first companies, GCCs, and Indian product-led companies. Available immediately. Open to relocation across Indian metros.
- 💼 LinkedIn — primary inbound
- 🌐 aksheywalia.in — portfolio + voice-AI digital twin
- 📧 aksheyw [at] gmail [dot] com
Last updated: 2026-05