Paloma wants your CRM to stop going quiet after "closed won."

In business-to-business software, everyone remembers the sale. The champagne is uncorked when a customer signs, the deal is logged, and the pipeline moves on to the next name. What happens next is where the money quietly leaks - and it is exactly where Paloma has decided to plant its flag. The San Francisco startup, part of Y Combinator's Summer 2025 batch, calls itself the system of record for "everything that happens after the handshake."

That phrase is doing a lot of work. Most customer relationship management tools - the Salesforces and HubSpots of the world - are built to win the deal. They track leads, log calls, and manage a pipeline toward a signature. After that signature, the same tools tend to go dark. The contract gets filed in a drive. Usage data lives in a warehouse. Billing happens in a spreadsheet three tabs deep, maintained by someone who dreads month-end. Paloma is a bet that the biggest revenue opportunity for most companies is not the next new logo. It is the customer they already have.

01 / THE PROBLEMThe revenue that leaks between contract and invoice

Ask a revenue leader where their growth actually comes from, and a growing number will say the same thing: expansion. Renewals and upsells inside the existing customer base are cheaper to win and more predictable than net-new deals. But the work of finding those moments is manual and easy to miss. An account manager is effectively asked to read a customer's mind - to notice that usage spiked, that a support thread turned tense, that a contract is quietly up for renewal in six weeks.

Meanwhile, the billing itself is its own tax on growth. Every custom term, every usage tier, every mid-cycle change means someone reads a contract by hand and translates it into an invoice. That translation is where errors live, and errors in billing are errors in revenue. Paloma's founders describe the result plainly: teams end up reactive, and growth gets constrained by operational bottlenecks rather than by how much a customer actually wants to buy.

The fragmentation has a cost that rarely shows up on a dashboard. When the contract sits in one system, the usage data in another, and the customer's last three complaints in a fourth, no single person holds the full picture. Renewals sneak up. A usage overage that should have triggered an upsell conversation instead becomes a surprise on an invoice, and a surprise on an invoice becomes a support ticket. The information exists - it is just scattered across enough tools that acting on it in time takes more coordination than most teams can spare.

02 / THE PRODUCTOne source of truth, then a copilot on top

Paloma works in two moves. First it consolidates the scattered evidence of a customer relationship - contracts, communications, notes, and usage data - and parses it into a single shared context. Revenue, finance and customer success teams read from the same page instead of three different ones. Second, it acts as an AI copilot on top of that context: flagging upsells, renewals, churn risks and usage changes, and ranking them by likely revenue impact so a team works the biggest opportunity first rather than the loudest email.

The most concrete piece is billing. Paloma reads the messy language of a signed contract and converts it into executable billing instructions - the part of revenue operations that almost nobody wants to own. It is designed to sit on top of the tools a company already uses, connecting to an existing CRM, billing software and data sources rather than demanding a rip-and-replace. Crucially, the founders say humans stay in control, with the system automating more only as it earns trust through feedback.

03 / THE FOUNDERSPeople who felt the pain before they built the fix

Founder-market fit is an overused phrase, but it is hard to avoid here. All three of Paloma's founders come from Deel, the global payroll and compliance company that became one of the fastest-growing startups of its generation. Nazli Danis, the CEO, and Alex Avnit, the CRO, spent three years building Deel's largest revenue product and watched it scale from zero to a reported $500 million in annual recurring revenue. Kaiwen Song, the CTO, led the development of Deel's billing and payments infrastructure - the exact machinery Paloma is now trying to make easier for everyone else.

The through-line is that they are not guessing at the problem. They lived through the version of it that only shows up at scale, when a revenue product is large enough that a small percentage of leakage becomes a real number. Building for the chaos you personally survived is a different exercise than building for a market you read about.

It also shapes what the product chooses to do and not do. A team that spent years inside billing infrastructure knows that the hard part is rarely the happy path - it is the edge cases, the amendments, the customer who negotiated a bespoke discount that nobody remembered to encode. Paloma's insistence on keeping a human in the loop, and on automating more only as trust builds, reads less like caution and more like scar tissue. People who have watched an automated system bill a customer wrong tend to design the next one carefully.

04 / THE MARKETWhere a post-sales CRM fits

The category Paloma is stepping into is crowded, but it is crowded in pieces. Traditional CRMs own the pipeline up to the sale. Customer success platforms like Gainsight watch for churn. Billing engines like Stripe Billing, Metronome and Zuora turn usage into invoices. Each solves a slice, and the seams between them are where Paloma's whole thesis lives. Its argument is that post-sales revenue is one continuous job - context, signal, and billing - that has been artificially split across systems that do not talk to each other.

There is a second, quieter story visible on Paloma's own website, which has also spoken about purpose-built software for "companies that make real things" - manufacturing and operations teams drowning in spreadsheets and manual data entry. Whether that reflects an early customer set, an experiment, or an evolving focus is not fully clear from public materials. What is consistent across everything the company has published is the underlying instinct: find the operational work a team would automate themselves if they had the engineers, and hand them the engineer.