How Cirrus Capital turned website chat into qualified deal flow and automated investor-to-company matching.
Challenge
Manual investor matching couldn't keep pace with growing deal flow
Cirrus Capital is a Florida-based debt capital advisor and co-investor. The business sits in the middle of a two-sided market: companies looking for debt capital on one side, investors with specific mandates on the other. Cirrus's job is to make matches that work for both parties — but the operational layer that supports those matches had been built piece by piece over the years, and it was holding the business back from the growth it had in front of it.
The first problem was the deal pipeline itself. Deals at Cirrus move through multiple stages — initial contact, qualification, structuring, investor introduction, term sheet, close. The team needed to know at any moment where every opportunity actually sat, who was responsible for the next step, and what had stalled. The existing setup couldn't deliver that. Pipeline updates happened in conversations, spreadsheets, and inboxes that were always slightly out of sync, which meant leadership made decisions without confident visibility and account owners reconstructed status manually every week.
The bigger structural problem was matching. Cirrus's value to a company is the right investor on the other end of the deal. Investors have explicit mandates — sector, deal size, geography, structure, stage — and companies have explicit needs. Pairing them required reading every investor profile against every active deal, and that was a job no one had the time to do consistently. Good fits got missed because the matching exercise lived in someone's memory rather than in any system. As deal volume grew, the gap between the matches Cirrus could find and the matches that actually existed got wider, not narrower.
On top of both, the website was generating real traffic that never converted into identifiable opportunities. Companies arrived looking for capital, browsed, and left — and Cirrus had no record of who they were, what they needed, or whether any of them were a fit for the investors already in the database. The top of the funnel was leaking the most qualified prospects the business was ever going to see, and there was no mechanism in place to catch them.
The compounding cost was structural. To support the deal volume Cirrus's growth plan called for, the only obvious answer with the existing setup was more people — more analysts to track pipeline, more advisors to run matching by hand, more sales staff to chase top-of-funnel leads. That's not the unit economics a debt capital advisory wants to scale on. Cirrus didn't need more headcount. It needed an operating system that did the work the headcount would have done.
Solution
A custom CRM, automated workflows, and a RAG-powered chatbot that pre-qualifies and matches deal flow in real time
A custom CRM built around the three databases that actually run the business
We started with the Catalytics Audit. Every workflow from website visit to investor introduction got mapped, and the data architecture was designed around what Cirrus actually moves through the business — not a generic sales pipeline. Most CRMs assume a 'lead, opportunity, deal' shape that has nothing to do with how a debt capital advisor thinks. Cirrus thinks in companies looking for capital, investors with mandates, and partner relationships that introduce both sides. Those three are different objects with different fields, different lifecycles, and different relationships.
We built a custom CRM with three distinct data foundations: a companies database, an investors database with structured mandate fields, and a partners database for the referral network. Each one was schema-designed for the information that actually drives matching and reporting. On top of that data, we built a custom CRM interface — drill-down records on every contact and every opportunity, deal-stage views the team could read at a glance, and integration with Beehive and SendGrid so email nurture sequences could trigger directly from CRM state. The result was the first version of Cirrus's operations that the team could actually see.
Automated workflows and a matching algorithm that runs every time either side changes
On top of the data foundation, we automated the work the team had been doing by hand. Email nurture sequences run through Beehive and SendGrid, triggered by deal stage and contact type — when a new opportunity hits a specific point in the pipeline, the appropriate sequence fires without anyone adding contacts to lists. Internal notifications fire when deals stall, when investor mandates change, and when new high-fit matches appear. The team stopped functioning as a manual data relay between systems and started functioning as decision-makers reviewing what the system had already done.
The matching algorithm is the engine inside it. It runs continuously against the structured investor mandate fields and the structured deal requirements — sector, ticket size, geography, structure, stage — and returns a ranked shortlist of fits whenever either side updates. What used to be a memory exercise across hundreds of investor profiles is now a query that runs in seconds. The team still validates the top matches with human judgment, but the search itself stopped being the bottleneck.
A RAG-powered website chatbot that turns website traffic into qualified, pre-matched deal flow
The AI layer is the conversion engine, and it's where the top-of-funnel problem actually got solved. We built a website chatbot using retrieval-augmented generation (RAG) over Cirrus's institutional knowledge — investor mandates, deal structures, sector expertise. A company arriving on the site is greeted by a conversation, not a contact form. The bot asks the right qualifying questions through natural language — what business, what stage, what kind of capital, what size, what timing — and the company fills in the information as a conversation rather than a five-field intake.
As that conversation progresses, the chatbot dynamically filters the investor database in real time. By the time the company has finished telling the bot what they need, a ranked shortlist of matched investors is already visible inside the chat. If the company wants to meet those investors, they're automatically added to an introduction email sequence and the new opportunity enters the CRM already qualified and pre-matched against specific counterparties. Website traffic that used to leave the site without a trace now leaves the site as an identified opportunity attached to a real shortlist.
"Our work with Catalytics will enable us to reach our 2024 and 2025 revenue and growth targets with 30–40% less headcount than anticipated."
Ryan Ridgway, Cirrus Capital
Results
Real-time investor matching, website traffic that converts into qualified pipeline, and growth targets reachable with 30–40% less headcount
Before the new system was live, Cirrus's team spent meaningful hours each week running matches by hand against an investor database that lived partly in spreadsheets and partly in partner-firm memory. They missed good fits routinely, not because anyone was careless, but because no human could keep every investor's full mandate held in mind against every active deal. After implementation, matching runs as an algorithm. Whenever a new opportunity enters the CRM, or an investor updates their mandate, the system surfaces the top fits automatically — ranked, scored against the criteria that actually matter, and visible to the team in real time.
Top-of-funnel changed completely. The Cirrus website used to generate traffic that produced no trace — visitors arrived, browsed, left, and Cirrus had no record of who they were. With the RAG chatbot live, that same traffic now converts into identified, pre-qualified opportunities that enter the CRM with their investor shortlist already attached. The conversion mechanism isn't a form anymore; it's a conversation that mirrors the kind of conversation an analyst would have had at intake — without anyone on the team having to take the call.
The biggest single change is structural. Cirrus's growth plan called for scaling headcount in line with the deal volume their pipeline was projected to generate. With the new operating system handling pipeline visibility, investor matching, lead qualification, and nurture sequences, that same growth path is reachable with materially less hiring than the original plan required. Ryan Ridgway, CEO, put it plainly: 30–40% less headcount than originally anticipated to hit the firm's 2024 and 2025 revenue and growth targets. That isn't a soft efficiency gain. That's the cost structure of the business changing shape — fewer hires needed, more deal volume supported per advisor, and a unit economics path Cirrus can actually scale on.
- 30–40% less projected headcount: same 2024–2025 growth targets, materially less hiring needed to support deal volume
- Real-time investor-to-company matching: deals pair against the investor database automatically as either side changes, surfacing a ranked shortlist instead of relying on manual matching
- Top-of-funnel converts: website traffic now leaves the site as identified, pre-qualified opportunities in the CRM with an investor shortlist already attached
- Single source of truth for the pipeline: drill-down records on every contact and opportunity, with email nurture sequences triggered automatically by CRM state
- A 24/7 deal-intake associate: the RAG-powered chatbot qualifies prospects through natural conversation, surfaces matches in real time, and routes qualified opportunities into automated investor introductions
"Our work with Catalytics will enable us to reach our 2024 and 2025 revenue and growth targets with 30–40% less headcount than anticipated."
Ryan Ridgway, Cirrus Capital
Let's get to work
Ready to build your Company Operating System?
Your business runs on 6 tools that don't talk to each other. We fix that.
