Product, Growth & AI · BrowserStack · ex-Amex · IIM Indore
I build tools that help people make better calls with their data.
I'm Samiksha — I lead product, growth, and AI at BrowserStack, where I've taken three AI products from zero to daily use. Before that, product analytics and customer strategy at American Express. I care less about the model and more about whether it actually changes a decision.
Things I've builtAll of them →
Self-Serve BI Platform
2025Let non-analysts ask the data questions themselves
ClaudeReactNode.jsBigQueryCloud Run
Read case study →Deal-Risk Detection
2025Catch deals slipping while there's still time to act
GeminiCall transcriptsNode.js
Read case study →Revenue Intelligence Copilot
2026A revenue assistant that can explain its own answers
PythonGeminiLangChainBigQueryReact
Read case study →How I think
The same idea runs under everything I build: the gap between data and decisions is a product problem.
- 01Start with the decision, not the dashboard. Most data work stalls at the last mile — turning a number into something someone acts on.
- 02Trust is the product; accuracy is just the price of entry. A correct answer nobody believes changes nothing.
- 03Good AI reduces what people have to hold in their heads. If it adds cognitive load, it's a worse product than the spreadsheet it replaced.
- 04The smallest change to a system usually beats the flashiest feature. I look for the leverage point, not the launch.
- 05If a model can't be questioned, it won't be trusted — and if it isn't trusted, it won't get used.
WritingAll essays →
Lately
- Building
- A revenue tool that explains its scores instead of just handing you a number.
- Figuring out
- How to test AI features when the "right" answer isn't a single fixed thing.
- Thinking about
- Why good dashboards so rarely turn into decisions anyone acts on.
Working on something hard in product or AI?
Those are my favorite conversations. Say hello.
Email me →