A few months ago, two people reached out to me on Upwork with a pretty bold idea.
Torrence Jennette, who’s a pro in the investment world, and Mallory Musante, an expert in marketing analytics, noticed something that most people in finance just accept. They realised that the way research is done hasn’t really changed in years, and it’s holding people back.
Right now, analysts at mid-sized funds are so buried in paperwork that they often miss great opportunities. There just aren’t enough hours in the day to manually dig through every filing and find those hidden gems.
Two months invested in audit and research
One thing I’m really proud of is that we didn’t rush to code anything right away.
Honestly, the biggest risk in tech isn’t that you can’t build something — it’s building something nobody actually needs. We spent two months in the audit phase specifically to make sure the development budget was invested where it creates value for the client.
By understanding how funds operate, we realised that for an analyst to win, they don’t need a generic AI. They need a system that mimics a high-performing research team. Luckily, Torrence and Mallory could share this exact knowledge, and our task was to make a context-aware AI system out of it. That’s when we decided to build a team of specialised AI agents: one for news, one for financials, one for valuations, all managed by an orchestrator.
We worked in two-week cycles, providing regular demos and gathering feedback. We linked four data sources, covering over 3,000 US companies — more than any analyst could cover manually. Next, we made natural language queries return real answers, added alerts, and built the rest of the functionality that matters to users.
I believe the key feature we worked on is coverage. Rather than processing isolated queries, the platform acts as a context-aware intelligence system. Analysts don’t start from scratch each time; they build on the knowledge they’ve already accumulated.
What Strem looks like today
Fast forward to now: Strem by BRDGE Insights is live, monitoring 3,000+ US securities 24/7. While an analyst might spend weeks digging through dry filings to find one hidden gem, the platform now handles that heavy lifting. It filters the noise so the investment team can focus straight on high-value insights.
The core mission behind Strem is to provide institutional-grade firepower. It allows mid-sized funds to compete with giants without the need for a massive research staff. With the foundation built, the project is already expanding into global markets and alternative data to catch market shifts before they even hit the news.
The case studies
We’ve written up the development process in two parts. The first covers the audit phase and the creation of the initial prototype. The second walks through the key features and explains how the platform accumulates and uses context.
Something not about AI
This year I joined a community of people passionate about business development, called Sales Label. I saw that Anton Fedulov, one of the founders, is actively developing this space and constantly shares insights that actually have value.
Since we were both living in Lisbon and on the same wavelength, I reached out to grab a coffee. We met up, and then for our next meeting we ended up doing a hike in Sintra. We spent the whole time talking about the housing market and what it’s like living in Portugal — basically the standard expat starter pack of topics. It’s cool how a simple message can turn into a genuine connection and a great Saturday afternoon.
Catch you in the next one,
Volodymyr
This issue first went out to subscribers of our LinkedIn newsletter, AI Era Development Stories, on June 4, 2026.
