Enterprise sales teams have spent years and millions of dollars deploying technology meant to make selling smarter. CRMs capture every interaction. Call recording tools transcribe hours of customer conversations. Marketing automation platforms track engagement across channels. Product analytics measure usage patterns down to the click.
Yet for all this investment, most sales representatives still walk into customer calls relying on gut instinct and generic playbooks. The data exists, but it sits fragmented across systems designed for managers to monitor performance rather than for frontline sellers to use in real-time. And when AI tools do attempt to surface insights, they often fail at accuracy. Representatives consistently report that these systems miss critical deal-moving details and lack the intuition that experienced sellers bring to complex conversations.
Overlayy Labs Inc., a B2B sales intelligence startup founded by Anmol Chaman and Snehil Saluja, set out to close that gap, and in doing so, they’re showing why so many AI-powered sales tools have failed to deliver on their promises.
The Problem: Signals That Never Reach the People Who Need Them
The disconnect between data collection and data utility in enterprise sales is decisively structural. Organizations pour resources into building technology stacks that generate vast quantities of information, but that information flows into dashboards built for leadership oversight rather than into the hands of representatives preparing for their next call.
“The GTM space is in data overload right now,” explains Anmol Chaman, co-founder and CEO of Overlayy. “You have a bunch of signals coming in, where any company has its CRM, marketing tools, and call transcription tools capturing a lot of data. But those tools don’t necessarily help take revenue forward.”
The result is a peculiar form of organizational amnesia. When a sales representative faces a difficult objection from a mid-market manufacturing prospect, there’s a chance that a colleague successfully dealt with the same objection six months earlier, but the approach they used to accomplish it is buried somewhere buried in a call transcript or scattered across email threads. Surfacing it, thus, requires manual searching that no one has time to do, so the representative improvises instead.
Playbooks created by sales leadership are meant to address this problem, but they’re often static documents that don’t reflect patterns emerging from actual customer conversations happening across the organization each week. And a significant category of valuable signals never enters the system at all: in-person meetings, hallway conversations, and offline interactions generate insights that remain locked in individual memories rather than contributing to collective intelligence.
Because of this, reports show sales representatives spend just 28% of their week actually selling, with the majority of their time consumed by deal management, data entry, and searching for information that should already be at their fingertips.
How Overlayy Solves This
Overlayy seeks to fix this by connecting to customers’ existing infrastructure (CRM systems, meeting note-takers, product analytics, marketing tools) to ingest historical deal data as well as ongoing communication streams into a unified layer. The platform then uses a proprietary time-series based context graph approach to map out patterns in historical deals: how similar customers were pitched, which case studies resonated, what objections arose, and how successful representatives handled them.
“We started connecting with our customers’ CRM, their meeting note-takers, their product analytics tools, and marketing tools to ingest historical data as well as ongoing deal data and create patterns out of it,” Saluja explained. “Our algorithms look at historical deals, look at what’s happening with the existing deal, and simply start creating patterns.”
The result is that, for example, when a representative prepares for a conversation with a particular type of customer (whether they work in a specific industry or have certain specific requests), Overlayy surfaces relevant context: metrics to emphasize, case studies to reference, likely objections to anticipate, and approaches that worked in comparable situations.
The system also captures signals from offline interactions, so representatives quickly log in-person meeting notes and critical context from non-digital conversations enters the deal record.
Why Getting It Wrong Is Worse Than Getting Nothing
The technical challenge that the Overlayy founders realized was essential to making the product actually helpful lay in making sure the system was as accurate as possible. Specifically, this meant the system had to not only not miss relevant information, but also not fabricate absent information.
The problem can be traced to how other competitors in the space handle long-context sales data: processing entire transcripts through single large language model calls can be expensive, but it can also lead an AI to misrepresent or create false information (known as “hallucinations”) if it’s not well-trained to associate the different signals, leading vendors to provide reports that lose critical details or, worse, aren’t rooted in reality.
For enterprise deployment, this can be particularly troublesome. Fabricated stakeholder names, invented budget figures, or made-up competitive mentions can lead sales teams to pursue strategies based on false premises, which, in the long run, may be worse than systems that simply acknowledge uncertainty.
“Whenever I speak to reps and VPs of Sales, they believe AI will give them the worst roadmap and action items,” Chaman explains. “So we realized: if you solve the accuracy layer, in other words, the deal-context layer, you can actually build a foundational insight layer that, when plugged in with LLMs, can be used to build AI agents for sales that aren’t just accurate but faster and more efficient.”
Customer feedback has consistently highlighted accuracy as Overlayy’s primary differentiator: the system captured details that other tools missed, which in turn gives users more thorough and relevant recommendations. For example, an early customer, a sales rep, pointed out that they use and trust Overlayy more than their CRM, one place to come before every call. Through proprietary benchmarking, Overlayy achieved 95-97% accuracy on sales conversation understanding, compared to the 60-70% the founders measured from existing market solutions.
Making Human Connection The Irreducible Element
The founders’ core thesis is that AI’s purpose is taking care of the time sales representatives spend on human relationship-building (the one dimension of enterprise sales that can’t be automated) while handling everything ancillary.
“One thing that’s been valuable since the beginning is human-to-human connection,” Chaman explains. “Our idea is to increase the time for a sales rep to build that connection, while letting AI handle everything else.”
Enterprise sales, particularly deals involving contracts worth half a million dollars or more, at the end of the day, need human-to-human connections to be successful. Buyers are evaluating the people they will work with as much as the products themselves, but representatives today spend more time navigating fragmented data across CRMs, call tools, and marketing platforms than actually building those relationships.
So as AI advances from basic text generation to more sophisticated tasks, the importance of human connection has only become more prescient. Anmol Chaman and Snehil Saluja built Overlayy on the premise that trust and relationship-building in high-stakes enterprise transactions will remain human-centered even as AI takes care of routinary research, proposal generation, and administrative tasks.
Their bet is that the winners in AI-saturated markets will not be the platforms with the most features, but those that take the issue of data overload and turn it into time for the human connections that close deals.
Disclaimer: GeekWire newsroom and editorial staff were not involved in the creation of this content..
