Day.ai

Open a typical CRM on a typical Tuesday and you will find a graveyard of good intentions: half-filled fields, deals last touched in March, notes that say "great call!" and nothing else. Day AI is betting that the salesperson should never have typed any of it.

Day.ai is a small company in Winchester, Massachusetts with an unfashionable obsession: the software category everyone loves to hate. Its product, CRMx, listens to the meetings, reads the emails, watches the threads, and writes the record itself. Ask it "which of our top accounts are showing early churn signals, and why?" and it answers in seconds - not after a week of analysts joining three tables together. The company is roughly 35 people, it has taken about $24 million from Sequoia and friends, and it opened to the public in February 2026 after quietly serving around 120 customers for more than a year. It is, in the founder's own phrasing, trying to be "the Cursor of CRM."

01 / WHO THEY ARE NOWThe company that wants to delete data entry

An AI-native CRM, launched into general availability in early 2026.

Today Day AI sells a single, slightly subversive idea: that a CRM should be built for the AI agents that will use it, not for the humans who currently feed it. The flagship is called CRMx, where the "x" stands for context. Instead of presenting an empty form, it ingests the raw material of a relationship - calls, inboxes, calendars, Slack - and reconstructs the narrative behind every account, including the part most systems throw away: the "why."

The pitch lands because the people hearing it have lived the alternative. Anyone who has run a pipeline knows the quiet tax of "updating the CRM," a task that ranks somewhere between expense reports and dental work. Day AI's wager is that the tax was never necessary - it was just the price of software that couldn't listen.

02 / THE PROBLEM THEY SAWA system of record that records nothing useful

Manual entry in, fragmented insight out.

Here is the tension that runs through everything Day AI does. The CRM promises to be the single source of truth about your customers. To deliver on that promise, it asks the busiest people in the company to stop selling and start typing. They mostly don't. So the source of truth fills with gaps, and the gaps get papered over with reports that take weeks to assemble and are stale the moment they ship.

Day AI calls this "a system of delay." The label is harsh and, irritatingly, accurate. Simple questions - which deals are at risk, which customers are drifting, why a renewal slipped - require pulling threads across multiple teams and tools. By the time the answer arrives, the moment to act on it has usually passed. The data exists; it is simply scattered, late, and missing its reasons.

03 / THE FOUNDERS' BETThe people who built the old CRM, building the next one

Christopher O'Donnell and Michael Pici, HubSpot alumni.

There is a certain nerve in spending a decade building one of the world's best-known CRMs and then setting out to make it obsolete. That is more or less the founding story. Christopher O'Donnell spent more than ten years at HubSpot, where he was tapped to build the company's CRM and sales tooling - the very products that helped take HubSpot public. His co-founder, Michael Pici, was a HubSpot sales and product leader. They started Day AI in the spring of 2023.

Their bet has a specific shape. They watched generative AI become genuinely good at reading messy human conversation at almost exactly the moment the CRM's core weakness - depending on humans to transcribe that conversation into fields - became fixable. The founders did not set out to add an AI feature to a CRM. They set out to ask what a CRM would look like if it had assumed AI from the first line of code. Sequoia found the answer convincing enough to lead a $4M seed in 2024 and a $20M Series A in February 2026, with partner Pat Grady joining the board.

04 / THE PRODUCTCRMx, and the assistant you get to name

Meeting assistant, CRM and knowledge base in one conversation.

In practice, CRMx behaves less like a database and more like a colleague with a perfect memory. It prepares for meetings, takes the notes, updates the records, manages the to-do list, and keeps the pipeline current - the unglamorous work you would assign to a chief of staff if you could afford one for every rep. Each customer even names their own assistant, which is a small touch that quietly reframes the thing from "feature" to "hire."

The interface is a conversation. Rather than navigating dashboards, you ask questions in plain English and CRMx reasons across the whole business to answer them. It folds together three tools that used to live in separate tabs - the meeting assistant, the CRM, and the knowledge base - into one surface. For a startup it can be the entire CRM; for a scaling company it can ride alongside whatever incumbent system already exists, which is a polite way of saying it does not demand a religious conversion on day one.

05 / THE PROOF120 customers and a very interested cap table

Evidence, with the usual caveats about a young company.

The strongest evidence so far is patience. Day AI worked with around 120 customers for more than a year before letting anyone sign up publicly - an unusually long private stretch in a category racing to ship "AI" onto every landing page. The restraint suggests a team more interested in whether the thing actually works than in how quickly it can be announced.

The second piece of evidence is the cap table. Sequoia led both rounds, which is rarer than it sounds, and the Series A drew in Greenoaks, Conviction, Sound Ventures and Permanent Capital. Pat Grady taking a board seat is the kind of signal investors send when they think a category - not just a company - is about to move. None of this guarantees success; plenty of well-funded CRMs have been quietly euthanized. But the people who write these checks have seen the inside of a lot of go-to-market software, and they are leaning in.

06 / THE MISSIONBringing the customer back into the room

From system of delay to system of memory.

Strip away the funding and the product specs and Day AI's mission is oddly human: to bring the voice of the customer into every decision. The company talks about restoring the clarity a founder has on day one - when you personally know every customer, every complaint, every reason a deal closed - and keeping that clarity intact as the company grows past the point where any one person can hold it all in their head.

That is the deeper claim hiding inside a sales-software pitch. The problem was never really data entry. It was that companies forget their customers as they scale, and the CRM was supposed to prevent that and didn't. Day AI's answer is to make the remembering automatic, so the knowledge compounds instead of evaporating with every rep who leaves.

07 / WHY IT MATTERS TOMORROWIf agents are coming, they will need a memory

A CRM designed for software that acts on its own.

The forward bet is simple to state and hard to dismiss. If autonomous agents are going to do real work inside companies - drafting the follow-up, flagging the risk, nudging the stalled deal - they will need somewhere to read the full context and somewhere to write back what they did. Today's CRMs, built around human forms, are a poor home for that. Day AI is arguing that the system of record has to be rebuilt for agents first, and that the company which owns that layer owns something valuable.

It is early, and the giants are not asleep; Salesforce and HubSpot can both spell "AI." Day AI's advantage is that it is not retrofitting - it started from the assumption that turned out to be true. Whether that is enough to unseat incumbents with decades of lock-in is the open question, and an honest reader should keep it open.

So return to that Tuesday morning CRM, the one full of blank fields and good intentions. Day AI's whole reason for existing is that you should never have opened it. The record should already be written, the context already there, the churn signal already flagged - and the only thing left for you to do is the part a human is actually good at: picking up the phone. The chore was never the job. Day AI is trying to prove it.