Explore my work.
Six companies. Different problems, products, and people. Select a company to explore the projects within.

AWS — Management Console
4 projects+
AWS — Management Console
4 projectsMy work in AWS Console Experiences connected enterprise customer discovery, shared platform ownership, and launches spanning the wider AWS organisation. I worked with 30+ Fortune 500 enterprise customers on the different needs of IT administrators and developers, then translated those needs into Console experiences, public APIs, and launch priorities.
User Experience Customization
How enterprise discovery shaped a new AWS service, and why I launched account colours before the broader customisation experience.
My role: I led customer discovery, feature prioritisation, and the phased launch of AWS User Experience Customization, working with engineering on the product experience and programmatic configuration.
100K+ accounts in 12 weeksView project →Multi-session Console
Launching multi-session across AWS, then improving service support, account identification, link sharing, and usability to reduce opt-outs.
My role: I led the launch, coordinating 250+ AWS teams. After launch, I combined user feedback with behaviour data, secured approval for an adoption roadmap, and drove delivery over six months, including follow-up work with 25+ teams on service support.
Opt-out rate: 20% → below 4%View project →Centralized Settings Platform
Owning shared settings infrastructure used by 100+ AWS services and storing 6B+ settings.
My role: I owned the centralised settings platform used by more than 100 AWS services. My scope combined platform ownership with new settings experiences and coordination across Console, service, and Identity teams.
Platform scope: 100+ services · 6B+ settingsView project →GenAI Launch Agents
Reusable AI agents for the repeated work around AWS Console product launches.
My role: I built reusable GenAI agents using Amazon Quick Suite to automate product-launch tasks for AWS Console PM teams.
~90% less time on launch tasksView project →
Jumbotail
2 projects+
Jumbotail
2 projectsI joined an early-stage B2B grocery marketplace and grew with it from hands-on Decision Sciences into team leadership and product management. I set up foundational data tools, wrote production Python algorithms for recommendations and bin packing, and built a five-person function. Later, I owned vendor experience and launched operational decision support across demand, inventory, pricing, logistics, and P&L.
Decision Sciences and Supply Chain
From the first data foundations to production ML and a five-person Decision Sciences function.
My role: I was the founding member of Decision Sciences, working solo for nearly a year before building and leading a five-person team. I also wrote production Python algorithms for the recommendation engine and bin packing.
5-person function · Production Python and ML systemsView project →Retailer Credit Rating
A credit-rating approach built around the business data of kirana retailers.
My role: I defined Jumbotail’s internal credit-rating system for kirana retailers, translating business and platform data into an approach for lending decisions.
Internal credit-rating system for kirana retailersView project →
Intello Labs
2 projects+
Intello Labs
2 projectsIntello Track brought computer vision into enterprise produce-quality workflows. My work covered the shared SaaS product, customer discovery and delivery, and the internal tools that helped models improve. I coordinated engineering, onboarding, data science, and QA, including customer engagements that delivered an initial app in roughly 4–6 weeks and led to contract renewals.
Produce Quality Assessment
Turning computer vision for produce quality into a shared enterprise SaaS product.
My role: I led the product’s global launch and the transition from bespoke customer apps to a shared SaaS offering. I also owned enterprise product engagements, coordinating engineering, onboarding, data science, and QA.
15+ enterprise customers · 7 countriesView project →Data Labeling and Model Iteration
An internal annotation tool to help the ML team improve models faster.
My role: I built an internal data-labeling tool to support annotation and model iteration for Intello’s produce-quality models.
Faster annotation and model iterationView project →
CRED — CRED Mint
1 project+
CRED — CRED Mint
1 projectCRED Mint was CRED’s first investment product, backed by peer-to-peer lending. I owned the online and offline product experience to build trust: how members understood the offering, moved through the experience, and received support. My scope was the trust and customer experience, alongside the launch strategy.

Maitri.life
1 project+
Maitri.life
1 projectMaitri.life was an entrepreneurial attempt to make everyday habits easier to sustain through community and accountability. I co-founded the venture, recruited an eight-person team, and launched a device-connected experience across mobile and WhatsApp for steps, meditation, and workout goals. We attracted real users, ran a three-month program, and later discontinued the venture when I pursued my MBA.

Credit Suisse — Prime Services
2 projects+
Credit Suisse — Prime Services
2 projectsI started in Prime Services, working on client pricing and strategy for hedge fund clients. My work combined financial analysis with tools that made it easier to act on that analysis: a client-level balance-sheet allocation methodology and a unified client-performance dashboard. The pricing work informed deal renegotiations; the dashboard later became part of the external client portal.
Balance Sheet Allocation and Pricing
A client-level balance-sheet allocation methodology to inform hedge-fund financing and pricing.
My role: I designed a new client-level balance-sheet allocation methodology that enabled more accurate return-on-assets analysis and informed deal renegotiations.
~15 bps improvement in RoA · ~$4M incremental annual revenueView project →Client Intelligence
A single view of client performance that replaced hours of manual data synthesis.
My role: I built a Tableau dashboard drawing data directly from MySQL to bring client-performance information into one view.
~400 hours of manual analysis saved per monthView project →