Tech · AI · Builder
Hi, I'm Saquib. I like turning messy questions into things I can build, test, and learn from.
I'm a product manager at Capital One, working on the digital payment and payoff experiences for personal loans. When someone misses a payment because the flow confused them, or can't figure out how to pay off their balance early, it isn't just frustrating — it can quietly add to their interest, affect their credit score, or turn a routine financial moment into an anxious one. Most of my work is about making sure that doesn't happen.
I spent the first decade of my career in edtech — as a tutor, then in product at eGMAT, then co-founding an AI-powered test prep platform. Now I build small things on the side and write about what I learn during my quiet hours.
A few things I've been building.
AI Servicing Intelligence
Turns raw customer complaints and support transcripts into ranked product opportunities and PRD-ready briefs.
Clusters complaints, ranks themes by opportunity, and exports a PRD draft instead of a dashboard.
- AI workflow design
- Customer research
- Fintech
NextGen Capital RAG Intelligence
A retrieval system for answering questions over financial documents, built so retrieval quality and evaluation are first-class concerns.
Answers are grounded in cited sources, and quality is measured rather than assumed.
- RAG
- Evaluation
- Trustworthy AI
DocBridge AI
A pre-RAG normalization pipeline that cleans messy enterprise documents before anything reaches the vector database.
Documents are cleaned, confidence-scored, and routed before ingestion — not after hallucinations appear.
- Document AI
- Pipeline design
- RAG infrastructure
Notes I'm working through.
What a RAG Demo Hides: The Metrics That Actually Matter
A retrieval demo answering one good question proves almost nothing. The interesting part is everything the happy path never shows you.
Designing Evaluation Loops for AI Products
Evaluation is not a QA step at the end. For AI products it is part of the product, and treating it that way changes what you build.
Let AI Answer, or Keep It Human? A Decision I Got Wrong Once
At GMATWhiz we automated a piece of mentorship that should have stayed human. Here is the framework I wish I had used.