Outlit Wants to Be the Memory Layer for Your Customer Data

The Macro: Customer Data Is Everywhere and Nowhere

Every SaaS company has the same problem. Customer information is scattered across a dozen tools, and nobody has the full picture. Sales knows what’s in HubSpot. Engineering sees the PostHog data. Finance watches Stripe. Support reads the Slack channels and email threads. But ask a simple question like “Is this customer happy?” and you’ll get a different answer depending on who you ask and which tool they checked.

Customer data platforms were supposed to fix this. Segment (now owned by Twilio) pioneered the category by creating a unified event stream that feeds data to downstream tools. Hightouch and Census built reverse ETL products that sync warehouse data back to operational tools. Totango and Gainsight focused on customer success scoring. The market is large and growing, but there’s a gap. Most CDPs are infrastructure. They move data around. They don’t understand it.

The AI agent wave is making this gap more urgent. If you want an AI agent to handle customer interactions, it needs context. Not just “this person signed up on March 3rd” but “this person signed up, hit a bug on day two, emailed support about it, got a workaround, upgraded to the paid plan anyway, and just added three team members.” That kind of longitudinal, cross-tool understanding is what humans build intuitively over months of working with a customer. Machines need it handed to them in a structured format.

The Micro: One Record to Rule Them All

Outlit connects to Stripe, PostHog, HubSpot, Slack, email, and other tools, then unifies everything into a single customer record. You can query across every connected tool, and AI agents can monitor signals and act on them. The product is SOC 2 and ISO 27001 compliant, which matters because the companies most likely to pay for this are the ones most cautious about data security.

Josh Earle is the CEO and has spent over six years building SaaS products that collectively reached more than a million users. Leo Paz is the CTO with six-plus years building enterprise web applications, from no-code website builders to computer vision systems. They’re a two-person founding team out of YC’s Winter 2025 batch, based in San Francisco.

The “customer context for agents” positioning is deliberate and well-timed. Rather than competing head-on with Segment on data infrastructure or with Gainsight on customer success workflows, Outlit is positioning itself as the context layer that AI agents need to function. That’s a different buyer and a different value proposition. The person purchasing Outlit isn’t necessarily the data engineer who set up Segment. It’s the ops leader or product manager who wants their AI agent to actually know things about customers before taking action.

The query capability is the feature that I think separates this from a simple integration layer. Being able to ask “show me all customers who had a support conversation in the last week and increased their usage by 50%” across multiple tools without writing SQL or building a dashboard is genuinely useful. That’s the kind of question that currently takes a data analyst thirty minutes and a Looker query.

The Verdict

The timing is good. The market is moving toward AI agents that interact with customers, and those agents need context to not be terrible. Outlit is building the layer that provides that context. If the AI agent trend accelerates, and I think it will, the demand for products like Outlit grows proportionally.

My concern is defensibility. The integrations themselves aren’t hard to build. Connecting to Stripe’s API or pulling data from HubSpot is well-documented work. The value has to come from the unification logic, the query layer, and the agent-readiness of the output. If Outlit’s unified records are genuinely better than what you’d get from stitching together Segment events and warehouse queries, they have something. If it’s mostly a convenience wrapper, the larger CDPs will add this functionality within a year.

The SOC 2 and ISO 27001 compliance is a smart move for a startup this early. Enterprise buyers will ask for it, and having it already removes a common objection in the sales process. A lot of startups treat compliance as something to worry about later. Outlit having it now suggests they’re thinking about the sales cycle correctly.

At 30 days, the metric is integration depth. How many tools can they connect, and how clean is the unified record? At 60 days, it’s about whether AI agents built on top of Outlit data actually perform better than ones without it. At 90 days, I’d want to see expansion within accounts. If one team adopts Outlit and other teams at the same company start asking for access, that’s the signal that the product is delivering real value. The biggest risk is being too early. If enterprise AI agent adoption moves slower than expected, Outlit could be a solution waiting for its problem to fully materialize.