Lightfield Raises $47M Series A Led by a16z to Accelerate Growth

Lightfield, a startup building an AI-native customer relationship management platform, announced a $47 million Series A funding round led by Andreessen Horowitz on September 9, 2026. Co-founder Keith Peiris wrote that more than 5,000 companies have signed up for the product since its launch in November 2025.

Peiris said Lightfield has become the system of record for the fastest-growing companies and that teams are moving entire Salesforce implementations over to the platform. He wrote that the new capital will go toward accelerating the company’s growth and broadening the work its agent can take on for customers.

Andreessen Horowitz confirmed it led the round in a September 9, 2026 post by investors Joe Schmidt IV, Alex Rampell, and Fabrizio Serafini, in which the firm called Lightfield “the CRM for the agentic era.”

A CRM Built for Agents

In the announcement, Peiris argued that legacy CRMs such as Salesforce and HubSpot were architected more than twenty years ago for humans to update records and store data in fields, leaving agents without the context they need to produce reliable work. Lightfield, he wrote, was built from day one for agents, assembling a record that draws continuously from calls, emails, and meetings and holds enough context to surface which deals are at risk or what customers will want next.

That context, Peiris wrote, lets customers delegate real work to agents, including building pipeline from patterns found in closed-won deals, deploying agents that respond autonomously to customers, and onboarding new sellers with lessons from how their best deals were won.

Andreessen Horowitz’s post describes a flexible data model that needs no setup and assembles itself from a company’s emails and calls within minutes, resting on what the firm calls a temporal context graph that captures the reasoning behind each deal. The firm said Lightfield’s agents take on the updates, follow-ups, handoffs, and reporting that accumulate around the customer record, and that pricing follows the work performed and value delivered rather than seats purchased.

Lightfield’s product page says the platform connects to a company’s email, calendar, Slack, and LinkedIn accounts, captures customer interactions from those systems, and structures the record automatically. The page describes custom objects and relationships that reflect how a business operates, a temporal context graph that ties interactions together and retains the history of how records change, and an agent harness that combines a standardized SDK, a deterministic code sandbox, and continuous evals. Every record can be read and written through an API, MCP, and CLI, according to the page. The company also offers a white-glove migration service in which its agent ingests legacy CRM data and recreates the customer’s data model on day one, plus a two-way sync option for running Lightfield alongside Salesforce or HubSpot while workflows remain in place.

From Tome to Lightfield

Andreessen Horowitz said it has known founders Keith Peiris and Henri Liriani for years. According to the firm’s post, the pair previously built Tome, an AI presentation tool that grew to 25 million users, then concluded it was not the right strategy for building a massive company. They pivoted, cut the company down to a core group of engineers, and rebuilt around a problem they had experienced firsthand while scaling Tome: there was no AI-native CRM for growing companies.

The founders had laid out their case for the new company in “Why we built Lightfield”, an essay published November 13, 2025. It argues that legacy CRM tools fail founders at three stages: at day zero, when founders still searching for patterns get a system of empty fields; at first revenue, when manual data entry turns the CRM into a burden on learning; and at scale, when the record reflects only whatever employees happened to enter by hand. The essay says Lightfield is designed to preserve the raw traces of customer interactions, such as every email, call, and meeting, organizing them so that chronology, attribution, causality, and state are maintained, with summaries, fields, and dashboards treated as views derived from that record rather than the record itself. “We know what it feels like to scale a company to millions of users and then realize you’ve lost sight of what your customers actually want,” the founders wrote.

Andreessen Horowitz’s Investment Thesis

The firm said that since Lightfield’s launch late in 2025, thousands of companies have adopted the platform and teams have switched to it from Salesforce, HubSpot, and Attio. Its post framed the investment around a pattern in which each computing generation produces a new system of record, naming Siebel on client-server, Salesforce on cloud, and HubSpot on the inbound small-business segment. The AI-native generation of companies, the investors wrote, is being founded now and is choosing its system of record for the first time.

Peiris wrote that Lightfield’s mission is to give every employee and every agent at every company access to that business world model. Andreessen Horowitz noted that the company is hiring.