I lead product strategy, platform capabilities, and AI innovation for a global platform that has processed 2M+ applications from 1.5M+ students across 150+ countries — setting OKRs with executive leadership, coaching the product team as a player-coach, and working directly with a company co-founder on capital allocation and growth bets. And I still build: six products conceived beyond the roadmap, engineered end to end, all running in production today.
Before that: a decade of product in regulated financial services — at EXL, product owner of a core insurance policy-administration platform serving CNO Financial Group; at Capgemini, embedded at Manulife, leading enterprise security and financial-reporting programs. I started as an engineer, and I've stayed close to the data ever since.

Own product vision, growth strategy, and roadmaps for a global platform — leading product across Product, Design, and Data, and directing capital allocation against business outcomes. Launched Smart Profile and Smart Sync, the portfolio's biggest 0-to-1 build — students upload a document once, the platform reads it and extracts their information, and every document syncs automatically between their profile and all of their applications. No more uploading the same file application after application. Built the application-rematch engine: 6,900+ replacement applications auto-created — 16% of new volume — converting at 13% in its first 90 days.
The automation portfolio moves applications through pipeline stages without manual touchpoints — median processing time down 35% in six months, 40,000+ operations hours saved a year. Personally built five products in-role — DocuShield document fraud detection, UK Visa Interview Prep, Job Insights, an Application Assistant answering from live application data, and AB Product Brain, the intelligence engine that gives product, engineering, and leadership plain-English answers from every system in the company.
When the market contracted, the product org absorbed the work of the design, data, and program-management functions — I rebuilt how PMs operate so fewer people carry more scope: PMs design, query, and prototype themselves, on frameworks and tools I built and trained them on. Coach the PM team on portfolio strategy, OKRs, and resourcing, and run exit interviews with churned partners myself.
Set product OKRs with executive leadership and ran delivery across 15 engineering teams — the operating cadence of quarterly operational reviews and funding trade-offs that pointed engineering capacity at growth. Ran delivery cross-functionally across the organization — sales, CX, support, and operations behind one plan — through the platform's peak years: 750K+ applications created and 310K+ submitted in a single year, zero-downtime peak seasons.
Built the frontline feedback loops wiring customer support and sales insight into product priorities — lifting CSAT 25%. Worked directly with a company co-founder on capital allocation and growth bets.
Joined during hypergrowth to stabilize platform infrastructure as engineering scaling outpaced internal tooling. Built the AWS foundation, CI/CD pipelines, and release train end to end; onboarded 15 engineering teams; ran the cloud security programs that shut down bot and data-scraping attacks. Promoted to Director within 13 months.
Embedded at Manulife as Product Owner and technical SME: led the product rollout of an enterprise-wide multi-factor authentication program, modernized member statements and tax-slip reporting into automated, audit-compliant delivery, and translated regulatory mandates into platform capabilities engineering could build. Built the release governance that made delivery predictable in a regulated environment.
End-to-end product ownership of LifePRO — a core insurance policy administration platform — for CNO Financial Group: the complete cycle from client requirement to shipped feature to measured adoption. Took enhancements directly from client executives, converted them into prioritized roadmaps, and directed the 24-person cross-functional org that delivered them. Repositioned platform capabilities around how end users actually worked, and built the adoption dashboards that showed which features were used — evidence that steered every roadmap decision.
Delivered client platform integrations on LifePRO — the core policy administration system serving CNO Financial Group — end to end: scope, schedule, and the bridge between business stakeholders and engineering. Introduced structured sprint delivery into rigid waterfall handoffs, making releases predictable for a US insurer's core systems. Measured each release after it shipped against the value it promised the client, and fed that back into what got built next. Promoted to Program Manager with product ownership of the platform.
The bridge years: technical translator between business and engineering on the Wealth Management Accelerator platform. Designed feature expansions that let insurance clients launch new product lines; introduced agile cadences into waterfall delivery.
High-availability financial transaction systems. Redesigned legacy mainframe modules to cut production defects and hold up under peak processing windows.
Where it started: high-volume mainframe systems (COBOL/JCL) for enterprise financial clients. Client-facing from year one — gathering requirements, shipping production code, owning the fix when batch jobs broke at 2 a.m.
Delivering the portfolio is the job — the career section above is that story. These six are what I built on top of it: products I conceived, designed, and engineered myself, end to end.
CA firms in India run their clients' books in Tally and Busy — desktop software with no APIs — keying in thousands of entries a month by hand. SuperFile takes receipts as they already arrive, over WhatsApp or email, extracts and classifies each entry with AI, and posts it on confidence — no human in the loop — including an RPA bot that operates Busy's own import screens. Support is self-healing: an in-app report triggers a Claude-powered fix-bot that reproduces, fixes, tests, deploys, and replies. Live in production at a CA firm; every layer mine, including the engineering.
See it live: app.superca.in ↗Every detail of the platform in one queryable system — warehouse, codebase, tickets, documentation, analytics and monitoring events, live application data, partner feedback. A PM brainstorms a feature against it, a new developer finds exactly where a change belongs, leadership checks what the platform can do. When support says "this application won't open," one query traces the answer across events, live data, and code. 100+ verified feature specs, 100% product taxonomy coverage — analysis from days to minutes, adopted well beyond the product team.
Fraudulent documents are one of the biggest risks in international admissions. DocuShield takes every uploaded document apart — separating it into its underlying layers and inspecting each one for tampering. It goes as deep as re-computing the arithmetic on bank statements: if the numbers don't add up the way a genuine statement's must, it's flagged. Forged documents get caught before they reach an institution.
UK schools run credibility interviews, and students walk in unprepared. The student picks their target school; the tool serves the question set that school's interviews actually draw from, built from research into how each institution interviews. The student answers out loud. The tool evaluates each spoken answer, scores every part, and coaches question by question — closing with a scorecard, tailored suggestions, and a summary report.
Students pick programs with almost no visibility into what's on the other side. Job Insights maps every program on the platform to real labour-market data — matched through CIP codes to official Canadian sources — and answers the question that matters: for this program, how many jobs were posted in the last 60 days, and what salary can a graduate expect. Every program, mapped.
Most support tickets asked one question: what's happening with my application? This chat widget sits right on the application page, wired into every application API across the platform — so when a user asks, it answers from the live state of their own application: status, documents, what happens next. Built and productionized end to end, taking the questions that used to become support tickets.
SuperFile's login screen is as far as a visitor gets — the books behind it are real. Ask me for the walkthrough →
A bet earns engineering capacity when the case survives real data.
I build the working version myself and put it on the table.
No attachment — the capacity goes back to growth.
I still run churned-partner exit interviews myself.