Problem: Founders and creators wanted a real LinkedIn presence but could not sustain the writing cadence. Ghostwriters were expensive and generic AI drafts sounded like everyone else.
Approach: Built a purpose-specific AI writer that learns a voice from a handful of past posts, drafts new posts in that voice, and schedules them. Shipped v1 as a single-page app in under two weeks using agentic coding on a modern React stack.
Outcome: Live product accepting sign-ups. Draft-to-publish time dropped from an hour to under five minutes for early users. Actively iterating on voice-calibration.

Problem: Boarding kennels felt clinical and pet parents in Indian metros wanted trusted homes for their pets, not cages. Hosts existed on WhatsApp groups but there was no marketplace to find them.
Approach: Designed a two-sided marketplace with host verification, city-level search, and a booking flow tuned for mobile-first behaviour. Local SEO baked in at launch so city queries surface the platform.
Outcome: Hosts onboarded across multiple Indian cities. Marketplace live and taking bookings. Growth focused on tier-1 metros before widening.

Problem: Small businesses were paying for human support they could not scale and losing leads outside working hours. Off-the-shelf chatbots either could not answer or handed off badly.
Approach: Built an AI support layer that ingests a business's own knowledge base, answers within known scope, and hands off cleanly with full context when it cannot. Ships with a lightweight embed script, no plugin required.
Outcome: Deployed on live sites. Deflection rates measured per client rather than promised in a pitch deck. Iterating on the hand-off UX based on real transcripts.

Let's build something
worth shipping.

Tell me what you need. Real reply, usually within 2 hours.

Real products. Real users.