How Respan replaced its CRM with Octolane’s revenue superintelligence

Respan (YC W24, formerly Keywords AI) is the LLM observability and evaluation platform for AI agents. They process more than a billion logs and two trillion tokens every month, support millions of end users, and have grown 8x year-over-year.

When you grow that fast, two things break: your CRM data, and your tone.

For Respan, those problems were connected. Keeping a useful record of a customer conversation is only part of the job. The team also needs to act on that conversation, remember what was promised, and write a follow-up that feels like a continuation of the relationship.

The story of their switch to Octolane is about that whole workflow. Moving the records was the starting point. The more lasting change was having a CRM that helps carry the work forward after the data arrives.

The migration from Attio: 4 minutes, zero data loss

Respan had a lot in Attio. Years of accounts, contacts, deals, threads, and custom properties, data volume that usually makes a CRM migration a week-long project with a contractor.

They asked Octolane to do it. The whole thing took under 4 minutes.

"We had real data. Years of records, custom fields, conversation history. I expected something to break. Nothing did. Every record came over, every field was preserved, and Octolane enriched the data as it migrated."

Not a single record was missed. Octolane pulled the data, ran enrichment on every account and contact during the transfer, and self-organized everything into the right structure on the way in.

The important detail is what came across with those records. Custom properties and conversation history hold the context that makes a CRM useful. Preserving them meant Respan could bring its existing customer knowledge into Octolane, rather than treating the switch as a fresh start.

Enrichment happened during the transfer, too. The migration brought the records into Octolane and added information to the accounts and contacts as they moved. Organization was part of the same process, giving the team a structured starting point for the work ahead.

That matters because a fast import is only valuable if the resulting data is usable. In Respan’s account of the move, speed, completeness, and structure came together. The team did not have to choose which of those outcomes mattered most.

Never missing a follow-up

Respan sells to AI teams that move fast. A demo on Monday becomes a procurement conversation on Wednesday becomes silence by Friday if no one's paying attention.

"Octolane makes sure we never miss a follow-up. Every conversation gets tracked, every commitment gets surfaced, and the next step is drafted before I think about it."

A follow-up depends on more than knowing when the last email was sent. It depends on understanding what happened in the conversation and what someone committed to doing next. Respan’s description of Octolane puts those commitments at the center of the workflow.

Tracking the conversation, surfacing the commitment, and preparing the next message connect the customer record to an action. The team has a draft to work from while the context is still relevant, instead of relying on someone to remember every open thread.

For a team selling to other fast-moving AI teams, that continuity matters. A conversation can change direction quickly. Keeping the next step visible helps the relationship keep moving with it.

The part that's hard to fake: writing in your actual voice

This is where Octolane separated itself for the Respan team. It doesn't just suggest follow-ups, it writes them in the voice of the person sending them.

"It captures the smallest nuances. The way I open emails, the phrases I actually use, the tone I take with technical buyers versus founders. I read the drafts and they sound like me, not like a CRM trying to write an email."

The longer the team uses it, the sharper the voice gets. Every reply they send becomes a signal for the next draft.

The distinction is in the small choices. An opening line, a familiar phrase, or a change in tone can make a message feel personal. Respan specifically called out the difference between how they write to technical buyers and how they write to founders.

Those details are easy to lose when every message begins with a generic template. The team’s experience with Octolane is that the drafts reflect how the sender already communicates. The goal is to preserve a recognizable voice as the volume of conversations grows.

That makes the follow-up workflow more useful as a whole. Remembering to send a message solves one problem. Having a draft that sounds like the person sending it addresses another. Respan highlighted both as reasons the switch mattered.

Self-organizing as the team grows

At Respan's growth rate, organization is the silent killer. Stages drift, fields go stale, accounts get duplicated, and the data slowly stops being trustworthy.

"Octolane self-organizes. As we add accounts and run conversations, it keeps the structure tight on its own. I'm not maintaining the CRM. The CRM is maintaining itself."

This is the ongoing side of the migration story. Bringing clean, organized records into a new CRM is a good beginning. Keeping that structure useful as new accounts and conversations arrive is what makes the change last.

Respan described the familiar pressures of growth: drifting stages, stale fields, and duplicated accounts. These problems accumulate alongside the work of selling. If the structure becomes unreliable, the team has to spend more time maintaining the system it relies on.

With Octolane, the team describes organization as something that continues as they work. The customer record is part of an active workflow, alongside conversation tracking and drafting, rather than a separate administrative task waiting for attention.

The takeaway

Respan migrated years of Attio data in under 4 minutes with zero loss and full enrichment. Their team never misses a follow-up. Their outbound sounds like them, not like a template. And their CRM stays organized while they grow 8x.

That's what self-driving AI CRM means.