Ergo Wants to Be the AI That Actually Updates Your CRM

The Macro: CRM Hygiene Is Everyone’s Problem and Nobody’s Priority

Here’s a number that should bother every sales leader: Salesforce’s own research says reps spend only 28% of their time actually selling. The rest is admin work, internal meetings, and yes, updating the CRM. Forrester puts CRM adoption failure rates between 47% and 63%. That means more than half the companies paying for Salesforce or HubSpot aren’t getting the value because their teams won’t use the tools properly.

The problem isn’t the CRM software. HubSpot is genuinely good. Salesforce is powerful if you can stomach the complexity. The problem is the humans. Sales reps hate data entry. They close a call, they want to move to the next call, not spend ten minutes logging notes and updating deal stages. So the CRM rots. Pipeline data becomes unreliable. Forecasts are built on fiction.

This has created an entire category of CRM enrichment and automation tools. Gong and Chorus record calls and extract insights. Clari tries to fix forecasting. Scratchpad built a layer on top of Salesforce to make updates less painful. Dooly (now acquired by Notion) tried the same thing. People.ai tracks activity data automatically. The space is crowded, but the problem persists because no single tool has fully solved the gap between “what the rep did” and “what the CRM says the rep did.”

AI agents are the latest attempt. The pitch is simple: instead of making it easier for reps to update the CRM, take the rep out of the loop entirely. Let the AI watch the calls, read the emails, and update the CRM itself.

The Micro: Healthcare AI Founders Pivot to Sales

Ergo is built by Yash Dulla (CEO) and Ishan Sheth (CTO), both Georgia Tech graduates who previously co-founded Breezy Medical, an AI company focused on healthcare documentation. That background is directly relevant. Healthcare documentation has the same core challenge as CRM updates: professionals who hate paperwork, critical data that needs to be captured, and high costs when it’s done wrong. They’ve essentially taken the same thesis and applied it to a market with faster sales cycles and less regulatory friction.

Yash won the Stanford Hackathon in 2024. Ishan has ML and software experience across startups and financial institutions. They’re part of YC’s Winter 2025 batch, working with Jared Friedman. The team is small and hiring for operations, SDR, and growth roles, which suggests the product has enough traction to start building a go-to-market engine.

The product monitors customer communication across Zoom, email, and Slack, then keeps HubSpot and Salesforce up to date automatically. But Ergo goes beyond passive CRM updates. The “agents” part of “AI Agents for Revenue Teams” means the system also executes follow-ups. It watches for market signals (a prospect’s company announces a funding round, a champion changes jobs, a competitor gets bad press) and triggers mid-funnel outreach based on those signals combined with past interaction history.

That second piece is where it gets interesting and risky. Automated CRM updates are table stakes at this point. Automated outreach based on AI judgment is a much bigger bet. If the AI sends the right follow-up at the right time, it’s magic. If it sends a tone-deaf email after a prospect just laid off half their team, it’s a disaster.

The Verdict

I think Ergo is entering a crowded space with the right technical foundation. The Breezy Medical background is genuinely useful here. They’ve already built systems that watch professional conversations and extract structured data. Applying that to sales instead of healthcare is a lateral move, not a leap.

The risk is differentiation. Gong already does conversation intelligence. People.ai already does activity capture. Clari already does revenue intelligence. Salesforce and HubSpot are both building their own AI features aggressively. The “we do all of it in one tool” pitch sounds good but requires beating multiple incumbents simultaneously.

The automated outreach angle is the potential differentiator. If Ergo can reliably execute mid-funnel follow-ups that actually close deals, that’s a step function beyond what Gong or People.ai offer. Those tools tell you what happened. Ergo wants to do something about it.

In 30 days, I want to see CRM accuracy rates. What percentage of Ergo’s automated updates are correct without human correction? If it’s below 95%, reps will stop trusting it, and an untrusted CRM automation tool is worse than no automation at all. In 60 days, the question is whether the automated follow-ups are generating replies or getting ignored. In 90 days, I want to see pipeline velocity metrics: are deals moving faster for teams using Ergo versus their pre-Ergo baseline? The CRM update piece is solvable. The autonomous sales agent piece is where the upside and the risk both live.