Work Contact LinkedIn

MyMirro — CEO / Founder-Builder

Multi-product build across product, engineering, ops, and growth.

Problem

Build and scale consumer products with strong engagement and retention, under startup constraints and constant iteration.

Constraints

Lean team, fast cycles, shifting priorities, and the need to ship while learning directly from users.

What I did

  • Shipped 4 products end-to-end (strategy → execution → iteration).
  • Led a team of 8 across functions and delivery.
  • Owned the founder stack: product decisions, ops, finances, hiring, user conversations, networking, fundraising.
  • Built systems designed for rapid iteration and measurable learning loops.

Outcome

35K MAUs ~45% retention High engagement and repeated usage at scale Repeated MVP shipping in < 15 days when required

Key takeaway

Speed + ownership wins when you connect shipping to metrics, not opinions.

YOSO — Agentic Shopping Assistant + AI Search

Timeline: ~2 months • Role: Product + Execution lead

Problem

Help shoppers discover and decide faster using conversational guidance, improving conversion and basket quality.

Constraints

High SKU complexity, personalization needs, and the need for trustworthy responses grounded in catalogue + user context.

What I built

  • An agentic shopping assistant with access to catalogue, user inputs, and user information/context.
  • Training foundation using ~25 lakh purchase data and a fashion intelligence knowledge base built over time.
  • Conversation and search flows designed to reduce friction and guide intent → purchase.

Outcome

2.5× conversion increase +28% AOV

Why it worked

Because it combined high-signal data (purchase history) with real-time context, and shipped as a practical system—not a demo.

Perfora — Customer Support Agentic AI + Control System

Role: Builder / Product owner

Problem

Reduce repetitive support workload while improving speed and consistency of customer responses.

Constraints

Support workflows change frequently; the ops team needs control without engineering bottlenecks.

What I built

  • A customer support agentic AI for repetitive tickets and faster resolution.
  • A control system so the Perfora team could create custom workflows and deploy changes quickly.
  • An operational setup that made iteration safe and fast.

Outcome

50% reduction in manual tickets Significant cost savings and improved customer satisfaction

Key takeaway

Agents become valuable when the business team can control them safely and ship workflow changes instantly.

Beyond case studies

Other things I've owned end-to-end

Rapid MVP shipping systems
Team + delivery operating rhythm
User research loops → product iteration
Budgeting + financial ownership as founder
Partner/client-facing solution delivery