About — Rahul Bala

I'm a product manager for AI-assisted products. The part of the job I'm best at is deciding what an AI system is allowed to ship and where a human keeps the final call. By day I'm a PM on the Visual Experience product line (Studio, Live Preview, Visual Builder) at Contentstack, a composable DXP used by enterprise brands. On my own time I run an eval-gated LLM content operation as a one-person business, plus a few smaller products I build with AI coding tools to test ideas against real users.

The journey

I started in product operations at ShareChat, building the roadmap for Creator Hub and learning what consumer scale actually feels like. The lesson that stuck was that a faster feedback loop beats a cleverer feature most weeks. The creator feedback loop I set up cut query turnaround from seven days to two.

At Lokal I worked on products for Tier 2/3 India, redesigning onboarding (+9% completion) and share formats (+10% organic installs) from field usability sessions. Watching real users in their own context, not a lab, taught me to let observed behaviour overrule the roadmap.

Then three years as Senior Product Manager at Asianet News, steering the roadmap across the network's properties in eight languages. I rebuilt the editorial CMS from zero to one and integrated AI workflows: grounded quick drafts, cross-language translation, title and SEO suggestions, automated social creatives. The product call that mattered most was where the AI stopped. Every AI-assisted draft was approved by that language's editor before publish, which is why desks adopted the tools instead of resisting them. Editorial output tripled, and the work supported a period in which the network's users grew 130%.

Since 2025 I've been at Contentstack, where I took Studio, a composable visual page builder, from zero to one and drove its Early Access go-to-market, shipped Drafts & Auto Save across the suite, and built the company's first product-adoption baseline.

In parallel, I co-founded and bootstrapped TamilGlitz, an entertainment site I ran for seven years to 3M monthly users and $50K annual revenue. That's where I learned organic growth and content economics, and what it means to own every part of a product.

Independent builds

Since January 2025 I've run motomotar.comas a one-person AI content product; the case study covers the product decisions behind it. Around it I've shipped smaller products, each built end to end with AI coding tools to test one idea against real users: an Android app for a reality-TV voting audience in two languages, with10,000+ downloads and a 4.4-star rating, plus a companion web property that gets weekly updates and push notifications; a local commodity-price site that collects its data on a scheduled serverless job, live since September 2026; and a single-model car-price site built from a 197-page topical map, with 199 URLs live since 1 November 2026. A native iOS app is in progress for a December release. Each one has real users, which is the only way I've found to learn what AI-built products actually need.

What I believe about AI products

An AI product is an evaluation problem wearing a feature's clothes. If you can't score the output, you can't improve it, gate it, or trust it, and someone has to own where the bar sits. My own content operation publishes nothing that hasn't passed deterministic QA gates and an anchored LLM judge. That discipline came from being burned, not from a blog post.

Constrained generation with human review beats open generation almost everywhere trust matters. The newsroom systems that worked at Asianet generated drafts, suggestions, and structure, and left the publish decision with an editor. The systems that struggle are the ones that try to skip that boundary.

Language is product, not localization. Building for eight languages, and for users in Tier 2/3 India, taught me that vernacular users aren't an afterthought market. They punish products that treat translation as a checkbox.

Currently

I'm open to Senior Product Manager roles in AI products — especially applied AI, content and workflow systems, and consumer products. See/now for what I'm building and learning this month.

Get in touch

The fastest way is email: [email protected]. I'm also onLinkedIn andGitHub. My résumé ishere (with a PDF version). If you're hiring for AI product roles or want to talk shop about evals, content systems, or building for India, my inbox is open.