In November 2025, after roughly a year in stealth, a San Francisco company called Lightfield turned on its website and started letting founders log in [TechTimes]. The pitch is narrow on purpose: a customer relationship system that wires itself together from the email, calendar, and meeting data a young sales team already generates, and then suggests the next call, the next reminder, the next follow-up. The intended buyer is not a chief revenue officer at a Fortune 500. It is a four-person seed-stage team that has never bought a CRM before and dreads the idea of standing one up.
Before writing about a new tool, it helps to describe the standard of care. Today, most early-stage startups manage their pipeline through some combination of a shared spreadsheet, a Notion database, a free HubSpot seat, or, if a former enterprise seller is on the team, a stripped-down Salesforce instance that nobody updates. The work of logging calls, tagging contacts, and setting reminders falls on whichever founder last took the customer call. Conversations live in Gmail threads, Slack DMs, Zoom recordings, and the heads of two or three people. The first real CRM rollout typically happens somewhere between Series A and Series B, often painfully, and often after a deal has already been lost because nobody followed up. Lightfield is betting it can slide in well before that moment.
The bet
Lightfield sells what its investors describe as an "AI-native CRM that assembles itself from email, calendar, and meetings" [3 LSVP]. Co-founder and chief executive Keith Peiris has framed the go-to-market in blunt terms: reach companies "before HubSpot even thinks about you" [Contrary Research]. The wedge is early-stage venture-backed startups and vertical SaaS teams, the segment that traditional CRM vendors find too small to chase with a direct sales motion and too fragmented to win with self-serve alone. By capturing customer conversations automatically rather than asking a founder to type them in, Lightfield is trying to remove the single biggest reason early CRM deployments fail, which is that nobody has time to feed the system.
The product claims, drawn from the company's investors and from third-party trackers, point in the same direction: automated interaction capture, AI-suggested actions, and reminders that surface inside the tools a founder already lives in [Tracxn][Crunchbase]. None of this has been independently benchmarked yet, and Lightfield has not published customer case studies or retention figures. What it has published is a clear thesis about who it is for.
Why it could be big
The CRM category is enormous and durable, and the early-stage slice of it has been underserved for a decade. HubSpot built a public company partly by giving away starter seats to small teams and upselling them later, but its product was designed before large language models could reliably parse a sales call. A system that listens to meetings, reads email threads, and proposes the next action without manual data entry is a genuinely different shape of product, and it maps cleanly onto how a two-person founding team actually works.
The people behind Lightfield have done this kind of consumer-grade onboarding before. Peiris and his co-founders previously built Tome, the AI storytelling product that reached 25 million users and raised $43 million from a roster of AI-world investors [Forbes, 2023]. The pivot from Tome to Lightfield is itself part of the story: the team is reapplying what it learned about fast adoption loops to a category, sales software, where adoption has historically been the hardest part. Lightspeed Venture Partners is backing the new company, and Y Combinator appears in the founding lineage as well. Total disclosed funding stands at roughly $81 million across rounds, with a Series B on the books [PitchBook][3 LSVP].
| Signal | Detail |
|---|---|
| Total disclosed funding | ~$81M |
| Most recent round | Series B |
| Lead investor named | Lightspeed Venture Partners |
| Public launch | November 2025 |
| Prior company reach (Tome) | 25M users |
The team and traction
Lightfield was founded in 2020 and lists Keith Peiris, Henri L, Ves S, Pete Nichols, Jack Reed, and Matt Serna among its co-founders [Tracxn][LinkedIn]. Peiris is the chief executive [iHeart Boardroom Club Podcast]. The bench skews toward people who have shipped consumer-feeling software at scale, which is unusual in a category dominated by enterprise sales veterans. Public traction beyond the November 2025 launch is limited to what investors and the company have said about the product's design and target customer, and the company has not yet disclosed paying-customer counts or revenue.
The honest counterfactual
The sharpest question facing Lightfield is whether the early-stage segment can carry an $81 million venture bet. Bears will note that seed-stage startups are price-sensitive, churn frequently when they fail, and tend to graduate to whichever CRM their first VP of Sales prefers, which is often Salesforce or HubSpot [Contrary Research]. The bull answer, drawn from the same source, is that capturing a startup's conversational history from day one creates a switching cost that did not exist before: the data exhaust is the moat, and a team that grew up inside Lightfield has a real reason to keep its history there as it scales. Whether that retention story holds up at Series A and beyond is the central thing to watch.
What to watch
The next twelve months should answer three questions. First, whether Lightfield publishes any customer numbers, logo lists, or usage benchmarks that let outsiders gauge adoption beyond the launch press. Second, whether the company stays disciplined about the seed-stage wedge or starts chasing larger accounts where HubSpot and Salesforce will fight back hard. Third, whether the product can demonstrate that its AI-suggested actions actually move pipeline, the metric that early sales teams will judge it on. A Series C, or a meaningful expansion of the product into adjacent workflows like support or success, would be the clearest sign that the wedge is working.