The workflow before#
Bregal Sagemount is a growth-focused private equity firm, and it wins by seeing the right companies before everyone else does. Before this work, every prospect started with an analyst and a blank document. They read the company website, searched the news, pulled industry reports and pieced together funding history by hand. A first-pass profile took 4 to 6 hours. At that pace the team could research 20 to 30 companies a week and reached about 60% of the sectors it wanted to cover. The writing around each deal was manual too. Outreach emails, call notes and CIM summaries were all drafted from scratch, so the people best placed to judge a company spent much of their week assembling the facts about it.
What we built#
We built a deal sourcing system that does the first pass for the deal team. It watches the market, profiles each company it finds, scores it against Bregal Sagemount's investment criteria and sends the strongest fits to the team with a research package already attached. Analysts and partners still make every call. They just start each one with the homework done.
A market that watches itself
The system monitors company websites, news, industry reports and the firm's own internal documents, and keeps a profile for every company in the target sectors. When something changes, the profile changes with it. We put all of it on one governed data layer in Databricks so the deal team, the models and the reporting read from the same facts.
Fit scores that think like the firm
Each company gets a fit score against the firm's investment criteria, and the scores agree with analyst judgment more than 85% of the time. We fine-tuned models on Bregal Sagemount's past memos and deal evaluations instead of relying on a generic model, because what counts as a good fit is specific to this firm's thesis.
Research packages with sources attached
Every routed company arrives with a profile an analyst can act on: what the business does, how it has grown, where it fits and what to ask. Each claim links back to the document it came from. We built it that way so an associate can check a fact in one click instead of taking the model's word for it.
First drafts for the deal team
The same system drafts founder outreach, summarizes calls and CIMs and writes first drafts of IC memos. Everything is written back to Salesforce, where the deal team already works, so nobody has to learn a new tool to use it. People review every routed opportunity and every draft before it goes anywhere.
How we delivered it
We started with research and scoring, the part that took the most analyst time. Before routing anything live, we ran the scoring model against companies the team had already evaluated and tuned it until its calls matched theirs. Drafting for outreach, call notes and IC memos came next, rolled out to the deal team inside Salesforce. Model performance and drift are tracked, so the team can see when scores start to wander and retrain before they do.
Before and after#
| Measure | Before | After |
|---|---|---|
| Companies researched per week | 20 to 30 | 200+ |
| First-pass research per company | 4 to 6 hours | Under 30 minutes |
| Target market coverage | About 60% | 90%+ |
| Fit scoring | Analyst judgment, case by case | 85%+ agreement with analyst judgment |
What's next#
Next we're connecting sourcing to diligence prep and portfolio monitoring. When a portfolio company or a competitor shows a meaningful signal, it'll flow straight into the deal pipeline as a sourced opportunity with research attached. No one will need to notice it, write it up and hand it over first.
