About Me
Hi, I'm Balaji
I'm a Product Leader who spent the last year building, breaking, and rebuilding products at the frontier of AI tooling — not as a hobby, but as a deliberate exploration of what AI changes for product management.
I believe the most impactful PMs of the next decade will be the ones who can close the loop between product thinking and working software. AI is making that possible for the first time.
A Product Leader's AI Learning Journey
12 months. 7 stages of learning. 13 prototypes built. Exploring how AI reshapes what it means to build products.
The Story
12 Months of Learning
A deliberate journey through AI-native product development — from naive curiosity to structured mastery.
AI Exploration
Immersed in the AI landscape — studying foundation models, emerging tooling, and what AI-native product development could look like. Built intuitions, not apps.
AI Exploration
Immersed in the AI landscape — studying foundation models, emerging tooling, and what AI-native product development could look like. Built intuitions, not apps.
Vibe Coding
Jumped in head-first — prompting AI to write entire apps with minimal code knowledge. Shipped 7 prototypes in weeks. Learned what AI can do when you just describe what you want.
Vibe Coding
Jumped in head-first — prompting AI to write entire apps with minimal code knowledge. Shipped 7 prototypes in weeks. Learned what AI can do when you just describe what you want.
Structured Learning
Stepped back to understand the fundamentals — how LLMs work, prompt engineering patterns, RAG, agents, and evaluation. Turned intuitions into frameworks.
Structured Learning
Stepped back to understand the fundamentals — how LLMs work, prompt engineering patterns, RAG, agents, and evaluation. Turned intuitions into frameworks.
AI Orchestration
Moved from single-prompt apps to multi-step AI workflows. Built apps where AI agents collaborate, hand off context, and produce structured outputs reliably.
AI Orchestration
Moved from single-prompt apps to multi-step AI workflows. Built apps where AI agents collaborate, hand off context, and produce structured outputs reliably.
Spec-Driven Development
Discovered that writing detailed specs before prompting AI produces dramatically better results. The PM skill of clear requirements became a coding superpower.
Spec-Driven Development
Discovered that writing detailed specs before prompting AI produces dramatically better results. The PM skill of clear requirements became a coding superpower.
BMAD Method
Adopted the BMAD (Breakthrough Methodology for Agile Development) framework — structured AI-assisted development with clear roles, checkpoints, and review loops.
BMAD Method
Adopted the BMAD (Breakthrough Methodology for Agile Development) framework — structured AI-assisted development with clear roles, checkpoints, and review loops.
AI Agent Teams
The frontier — orchestrating specialized AI agents as a product team. Researcher, architect, developer, and QA agents working in concert to ship production-ready software.
AI Agent Teams
The frontier — orchestrating specialized AI agents as a product team. Researcher, architect, developer, and QA agents working in concert to ship production-ready software.
The Work
13 Prototypes Built
From quick vibe-coded experiments to structured AI-orchestrated products.
Aasaan
A simple prototype of a Product Management Learning platform built through pure prompting - no hand-written code.
Level Us Up
Gamified social consciousness app to shape the next generation.
Simple Banking
A clean personal finance dashboard for tracking income, expenses, and savings goals.
100 Products
A curated showcase of 100 micro-products built to explore different problem spaces.
Reaction Timer
Browser-based reaction speed test game — how fast are your reflexes?
Coffee Personality Quiz
Fun personality quiz that matches you to your ideal coffee drink based on your traits.
PingToProd
An AI-powered product leader who orchestrates a team of agents to research, design, build, secure, and ship your product.
Book Recommender
A book recommendation engine based on your personal preferences.
RoomieTab
Smart roommate expense splitter that uses AI to suggest fair divisions for shared bills.
EV Charge Smart
At a glance view of EV charging costs vs. gas cost across all 50 states.
Calorie & Sugar Tracker
Daily nutrition tracker focused on calorie and hidden sugar awareness for healthier habits.
Book Shelf Scanner
Point your camera at a bookshelf — AI agents identify, catalog, and rate every book.
Key Takeaways
What 12 Months Taught Me
Hard-won insights from shipping with AI every week.
Prompting is a Product Skill
Writing a great prompt is identical to writing a great PRD — clarity of outcome, constraints, and context. PMs who think in user stories and acceptance criteria have a natural edge with AI. The skill transfers directly.
Methodology Compounds Quality
Vibe coding ships fast but creates fragile apps. Spec-driven development and BMAD produce more maintainable results. Each methodology is a trade-off between speed and structural integrity — and knowing which to use is the real skill.
Orchestration is the New Architecture
Single-prompt apps hit a ceiling fast. The real leverage comes from multi-agent workflows: specialized agents with clear roles, handoffs, and verification loops. This is software architecture applied to AI systems.
PMs Should Own the Build Loop
When PMs can ship prototypes directly, the feedback loop collapses from weeks to hours. You stop translating ideas through multiple people and start validating directly. AI removes the translation tax between product thinking and working software.
About
Hi, I'm Balaji
I'm a Product Leader who spent the last year building, breaking, and rebuilding products at the frontier of AI tooling. Not as a hobby — as a deliberate exploration of what AI changes for product management.
I believe the most impactful PMs of the next decade will be the ones who can close the loop between product thinking and working software. AI is making that possible for the first time.
This site documents the journey — seven stages of learning, thirteen prototypes built, and the hard-won lessons that came from doing the work myself.