I'm Shayne, and these days I'm building review.ai — we're using AI to help teams turn their questions into executable knowledge. I've been lucky to work with some incredible teams over the years (was the first engineer at Instagram, worked on React Native at Meta, and more recently led engineering/product at Tailscale), and those experiences got me thinking about how we can make institutional knowledge more discoverable and actionable.

I'm a product engineer at heart — someone who loves diving deep into both the technical implementation and product thinking behind great software. Along with building products, I love exploring different passions — I'm a licensed pilot and once took a detour from tech to open a craft coffee shop. Life's too short not to pursue the things that spark your curiosity.

Right now I'm obsessed with transforming how teams understand their own data and decisions. We started with AI-augmented code reviews — giving engineers fast feedback while writing code. Then we built a research agent that could surface insights about your product, spanning from CRM data all the way down to your codebase. But that's when we realized the real opportunity: executable context.

Here's the thing — today's AI is designed for chat, not computation. Most tools either scrape random websites or dump raw API data into LLM context until it explodes. That's not analysis — that's expensive pattern matching. The interesting insights are hidden in your CRM, Gong calls, support tickets, user behavior data. All the stuff that actually tells you what's happening in your business.

So we're building something different: an answer agent that writes small Python programs to fetch, analyze, and understand your data on demand. Instead of dashboards that show you what happened, we generate executable code that helps you understand why it happened and what to do next.

Our approach turns every question into reusable infrastructure:

• "Categorize our support tickets by customer size and revenue to understand escalation patterns" becomes a program you can run, modify, and share
• Analysis pipelines become institutional knowledge that compounds over time
• Questions get answered with transparent, auditable code — no black box AI
• Teams build up executable understanding of their business, not just chat history

We're building towards a future where business intelligence isn't static dashboards, but living programs that grow smarter as your organization does.

We're bringing on our first design partners to help shape how this works in practice. If you're interested in turning your team's questions into institutional knowledge that scales, or just want to geek out about building the future of business intelligence (or swap coffee brewing tips!), drop me a line at team@review.ai.

Most of my commits these days are in private repos (startup life!)