Lightfield launches AI-native CRM after Tome pivot

2 days ago7 min read4 comments

Lightfield, a customer relationship management platform constructed entirely around artificial intelligence, has officially launched to the public following a year of stealth development—a remarkably bold pivot by a startup that previously operated as Tome, which had amassed 20 million users and secured $43 million in funding for an entirely different product. This San Francisco-based company is positioning itself as a fundamental reimagining of how businesses track and manage customer relationships, systematically abandoning the manual data entry that has defined CRMs for decades in favor of an architecture that automatically captures, organizes, and acts upon customer interactions.With over 100 early customers already integrated into the platform daily—more than half spending over an hour per day within the system—Lightfield represents a direct architectural and philosophical challenge to the legacy business models of Salesforce and HubSpot, both of which generate billions in annual revenue yet suffer from notoriously low user satisfaction. Keith Peiris, Lightfield's co-founder and CEO, articulated the core problem in an exclusive interview: 'The CRM, categorically, is perhaps the most complex and lowest satisfaction piece of software on Earth.CRM companies have tens of millions of users, and you'd be hard-pressed to find a single one who actually loves the product. That problem is our opportunity.' The general availability announcement marks a significant inflection point in enterprise software evolution, representing a company betting that large language models have advanced sufficiently to replace structured databases as the foundation of business-critical systems—a wager that attracted continued backing from Coatue Management, which led the company's Series A when it was still building presentation software. The story behind Lightfield's creation reflects both technical conviction and market pragmatism.Tome had achieved substantial viral success as an AI-powered presentation platform, gaining millions of users who appreciated its visual design and generative capabilities. However, Peiris explained that the team concluded building lasting differentiation in the general-purpose presentation market would prove difficult, even with a working product and genuine traction.'Tome went viral as an AI slides product, and it was visually delightful and easy to use—the first real generative AI-based presentation platform,' Peiris noted. 'But the more people used it, the more I realized that to truly help people communicate something—anything—we needed more context.' This realization prompted a fundamental rethinking, observing that the most effective communication requires deep understanding of relationships, company dynamics, and ongoing conversations—context that exists most richly in sales and customer-facing roles. Rather than continuing as a horizontal tool for everyone, they deliberately built vertically for go-to-market teams.'We chose this lane, 'sales,' because so many people in these roles used Tome, and it seemed like the most logical place to go vertical,' Peiris stated. The team reduced headcount to a core engineering group and spent a year building in stealth mode.Dan Rose, a senior advisor at Coatue who led the original investment in Tome, said the pivot validated his conviction in the founding team's capabilities: 'It takes real guts to pivot, and even more so when the original product is working. They shrunk the team down to a core group of engineers and got to work building Lightfield.This was not an easy product to build; it is extremely complex under the hood. ' What distinguishes Lightfield architecturally from traditional CRMs is foundational, not merely cosmetic.While Salesforce, HubSpot, and their competitors require users to define rigid data schemas upfront—dropdown menus, custom fields, checkbox categories—and then manually populate those fields after every interaction, Lightfield stores the complete, unstructured record of what customers actually say and do. 'Traditional CRMs force every interaction through predefined fields—they're compressing rich, nuanced customer conversations into structured database entries,' Peiris explained.'We store customer data in its raw, lossless form. That means we're capturing significantly more detail and context than a traditional CRM ever could.' In practical implementation, this means the system automatically records and transcribes sales calls, ingests emails, monitors product usage, and maintains what the company terms a 'relationship timeline'—a complete chronological record of every touchpoint between a company and its customers. AI models then extract structured information from this raw data on demand, allowing companies to reorganize their data model without manual rework.'If you realize you need different fields or want to reorganize your schema entirely, the system can remap and refill itself automatically,' Peiris elaborated. 'You're not locked into decisions you made on day one when you barely understood your sales process.' The system also generates meeting preparation briefs, drafts follow-up emails based on conversation context, and can be queried in natural language—capabilities that represent a fundamental departure from the passive database model that has defined CRMs since the category's inception in the 1980s. Early customer testimonials suggest the automation delivers measurable impact, particularly for small teams without dedicated sales operations staff.Tyler Postle, co-founder of Voker. ai, reported that Lightfield's AI agent helped him revive more than 40 stalled opportunities in a single two-hour session—leads he had neglected for six months while using HubSpot.'Within 2 days, 10 of those were revived and became active opportunities that moved to proof-of-concept,' Postle stated. 'The problem was, instead of being a tool of action and auto-tracking—HubSpot was a tool where I had to do the work to record customer conversations.Using HubSpot I was a data hygienist. Using Lightfield, I'm a closer.' Postle noted his response times to prospects improved from weeks or months to one or two days, a change noticeable enough that customers commented on the increased responsiveness. Radu Spineanu, co-founder of Humble Ops, highlighted a specific feature addressing what he views as the primary cause of lost deals: simple neglect.'The killer feature is asking 'who haven't I followed up with?'' Spineanu said. 'Most deals die from neglect, not rejection.Lightfield catches these dropped threads and can draft and send the follow-up immediately. That's prevented at least three deals from going cold this quarter.' Spineanu had evaluated competing modern CRMs including Attio and Clay before selecting Lightfield, dismissing Salesforce and HubSpot as 'built for a different era,' noting those platforms assume companies have dedicated operations teams to configure workflows and maintain data quality—resources most early-stage companies lack. Peiris claims the current batch of Y Combinator startups—widely viewed as a bellwether for early-stage company behavior—have largely rejected both Salesforce and HubSpot.'If you were to poll a random sampling of current YC startups and ask whether they're using Salesforce or HubSpot, the overwhelming answer would be 'no,'' he asserted. 'Salesforce is too expensive, too complex to set up, and frankly doesn't do enough to justify the investment for an early-stage company.' According to Peiris, most startups begin with spreadsheets and eventually graduate to a first CRM—a transition point where Lightfield aims to intercede. 'Increasingly, they're choosing Lightfield instead and skipping that intermediate step entirely,' he said.This represents a familiar pattern in enterprise software disruption: a new generation of companies forming habits around different tools, creating an opening for challengers to establish themselves before businesses grow large enough to face pressure toward industry-standard platforms. The company's strategy appears deliberately targeted at this window, aiming to grow alongside early customers and become embedded in their processes as they scale.Both Salesforce and HubSpot have announced AI features in recent quarters, adding capabilities like conversation intelligence and automated data entry to their existing platforms. The critical question facing Lightfield is whether established vendors can incorporate similar capabilities—leveraging their existing customer bases and integration ecosystems—or whether fundamental architectural differences create a genuine technological moat.Peiris argues the latter. 'The fundamental difference is in how we store data,' he emphasized.'Because we have access to that complete context, the analysis we provide and the work we generate tends to be substantially higher quality than tools built on top of traditional database structures. ' Existing conversation intelligence tools like Gong and Revenue.io, which analyze sales calls and provide coaching insights, already serve similar functions but require Salesforce instances to operate. Peiris contends Lightfield's advantage comes from unifying the entire data model rather than layering analysis on top of fragmented systems.'We have a more complete picture of each customer because we integrate company knowledge, communication sync, product analytics, and full CRM detail all in one place,' he explained. 'That unified context means the work being generated in Lightfield—whether it's analysis, follow-ups, or insights—tends to be significantly higher quality.' However, this architecture creates obvious technological and ethical challenges. Storing complete conversation histories raises significant privacy concerns, and relying on large language models to extract and interpret information introduces the possibility of errors—what AI researchers term hallucinations.Peiris acknowledged both issues directly. Regarding privacy, the company maintains that call recording follows standard practices with visible notifications, and that storing sales correspondence mirrors what CRM vendors have done for decades.The company has achieved SOC 2 Type I certification and is pursuing both SOC 2 Type II and HIPAA compliance. 'We don't train models on customer data, period,' Peiris stated.On accuracy, he was similarly forthright: 'Of course it happens. It's impossible to completely eliminate hallucinations when working with large language models.' The company's approach requires human approval before sending customer communications or updating critical fields—positioning the system as augmentation rather than full automation. 'We're building a tool that amplifies human judgment, not one that pretends to replace it entirely,' Peiris said.This represents a more cautious stance than some AI-native software companies have taken, reflecting both technical realism about current model capabilities and potential liability concerns around customer-facing mistakes. Lightfield's pricing strategy reflects a broader thesis about enterprise software economics.Rather than charging per-seat fees for a point solution, the company positions itself as a consolidated platform that can replace multiple specialized tools—sales engagement platforms, conversation intelligence systems, meeting assistants, and the CRM itself. 'The real problem is that running a modern go-to-market function requires cobbling together 10 different independent point solutions,' Peiris observed.'When you pay for 10 separate seat licenses, you're essentially paying 10 different companies to solve the same foundational problems over and over again. ' The company operates primarily through self-service signup rather than enterprise sales teams, which Peiris argues allows for lower pricing while maintaining margins—a common playbook among modern SaaS companies but representing a fundamental difference from Salesforce's model, which relies heavily on direct sales and customer success teams.Whether this approach can support a sustainable business at scale remains unproven. The company's current customer base skews heavily toward early-stage startups—more than 100 Y Combinator companies, according to the company—a segment with limited budgets and high failure rates.But Lightfield is betting it can become the system of record for a cohort of fast-growing companies, eventually creating an installed base comparable to how Salesforce established itself decades ago. The company's trajectory will likely depend on whether AI capabilities alone provide sufficient architectural differentiation—or whether incumbents can adapt quickly enough to defend their positions.The company has outlined several areas for expansion, including an open platform for workflows and webhooks that would allow third-party integrations. Early customers have specifically requested connections with tools like Apollo for prospecting and Slack for team communication—gaps that Postle acknowledged but dismissed as temporary.'The fact that HubSpot and Salesforce have these integrations already isn't a moat,' Postle contended. 'HubSpot and Salesforce are going to lose to Lightfield because they aren't AI native, no matter how much they try to pretend to be.' Rose highlighted an unusual use case that emerged during Lightfield's own development: the company's product team used the CRM itself to analyze customer conversations and identify feature requests. 'In this sense, Lightfield is more than just a sales database; it's a customer intelligence layer,' Rose noted.This suggests potential applications beyond traditional sales workflows, positioning the system as infrastructure for any function requiring deep customer understanding—product development, customer success, even marketing strategy. For now, the company remains focused on proving the core value proposition with early-stage companies.But the broader question Lightfield raises extends beyond CRM software specifically: whether AI capabilities have advanced sufficiently to replace structured databases as the foundation of enterprise systems, or whether the current generation of large language models remains too unreliable for business-critical functions. The answer will likely emerge not from technical benchmarks but from customer behavior—whether sales teams actually trust AI-generated insights enough to base decisions on them, and whether the efficiency gains justify the inherent unpredictability of working with systems that approximate rather than calculate.Lightfield is betting that the trade-off has already shifted in favor of approximation, at least for the millions of salespeople who currently view their CRM as an obstacle rather than an asset. Whether that bet proves correct will help define the next generation of enterprise software architecture and establish whether AI-native systems can truly displace decades of database-centric design patterns.

#featured

#Lightfield

#AI-native CRM

#Tome pivot

#Salesforce competitor

#enterprise software

#generative AI

Stay Informed. Act Smarter.

Get weekly highlights, major headlines, and expert insights — then put your knowledge to work in our live prediction markets.