Work
Experience
Suppliers emailed their updates to admins, who entered them by hand - 30–40 change requests a week, 2–4 hours per admin, and supplier data that advisors sold from was stale or wrong.
- Led the Supplier Portal for Brownell from discovery to launch, replacing emailed supplier updates that often got missed with a self-service portal, refined over 3 beta cycles with 10 suppliers and 2 travel agency admins.
- Launched the portal to 221 suppliers during peak travel season. Within the first month, 156 logged in and 78 submitted 128 updates, with 88% approved.
- Designed a configurable field model (DMC, hotel, cruise, villa, rep + custom fields) that replaced a single free-text field, so admins set up a new supplier type in under 5 minutes without developers.
Returned in May 2026 as Product Manager Associate (working student) after completing my internship.
Field knowledge from familiarisation (FAM) trips and conventions got buried in long Slack threads; marketing emails were the same for every traveler because HubSpot didn't know what they had booked or preferred.
Worked on two AI products that build on Trip Boutique's inventory and traveler data: a reporting workflow that turns advisors' field notes into structured reports, and the AI marketing module that matches travelers to trips.
Impact
- Owned an AI reporting workflow that turns advisors' audio and text notes into structured reports - 100+ field reports from FAM trips and conventions within 3 months, linked to 400+ hotels and restaurants.
- Co-owned the AI marketing module, leading its targeting and exclusion rules; in a Twerenbold pilot, revenue per campaign was 4× the prior year with 48% fewer unsubscribes, and the group is now expanding it to 4 more brands.
- Owned the module's second use case end to end - targeted campaigns to fill pre-planned group trips; 2 of 2 trips filled within 1–2 weeks.
- Contributed to onboarding of 3 enterprise customers during ARR growth from USD 80k to 210k.
Job discovery for blue and grey-collar workers in Tier-2 cities was fragmented and unstructured, validated by surveying 2,100+ job seekers and interviewing 18 companies.
Built a two-sided job marketplace from scratch. Grew to 6,000+ job seekers and 240+ employers across 2 cities, enabling 8,200+ job applications within 6 months of launch.
Cold-start → scale · hover a bar for detail
The 18 companies I'd interviewed became the first customers - solving the cold-start before launch, so supply and demand grew together.
- Built 2 products in 8 months (employee-appreciation software and WorkNote, a ticketing tool for car service centers). Both stalled because the solution was validated, not the problem; pivoted to Gowrk with a problem-first approach.
- Launched Gowrk, a two-sided job marketplace, scaling to 6,000+ job seekers and 240+ employers across 2 cities, enabling 8,200+ job applications within 6 months of launch.
- Solved the cold-start problem by converting 18 design partners into the first 20 job postings in week one, so job seekers found live listings at launch.
- Diagnosed low-quality applications as the top churn driver, then cut them by 40% with qualification-based matching, and added employer verification to protect job seekers from fake postings.
- Identified WhatsApp as the primary job-sharing channel via user interviews; launched sharing features, driving significant growth in new user sign-ups.
- Shipped an Android app for the job seeker side that reached 2,000+ active devices with a 4.9★ rating, mostly through in-app sharing.
Manual ERP onboarding and travel & expense workflows were slow and error-prone, creating multi-day resolution backlogs.
I automated ERP onboarding and the travel & expense workflows end to end, and managed the engineering delivery to get it done. Each onboarding batch went from 14 hours to about 4, and 93% of travel & expense tickets started closing the same day instead of in 5 days.
Slow → near-instant
- Led the automation of onboarding new users into PeopleSoft ERP in batches of 3,000–5,000, guiding 2 engineers and cutting each batch from 14 hours (manual) to about 4 (semi-automated).
- Automated travel and expense workflows, enabling self-service submissions and approvals and increasing same-day ticket resolution to 93% (vs. 5-day SLA).
Skills
Product
RICE Prioritization· Roadmapping· OKRs· Product Discovery· User Research· Go-to-Market· Agile (Scrum)
AI & Data
Generative AI· Prompt Engineering· SQL· KPI Frameworks· Human-in-the-loop Design· Data-driven Decision Making
Growth
Two-sided Marketplace· Cold Start· Demand & Supply Growth· Conversion Optimization
Leadership
Stakeholder Management· Cross-functional Collaboration· Influence Without Authority
Tools
Jira· ClickUp· Figma· Slack· MS Office· LogRocket
Education
Led a team of 5 to pitch an AI-powered ambient clinical intelligence SaaS for India - market sizing, competitive analysis, business model. Grade: 5.5/6.0.
Credentials
Languages