We Built This For Ourselves First

A personal experiment in radical self-observation became the most honest operational intelligence tool we'd ever seen.

The Origin

Reliquary started as one person's experiment in radical self-observation.

The founder built something he called Witness: 18 AI agents with permission to watch everything—calendar, email, code commits, browser history, health data. The instruction was simple:

Don't be nice. Tell me what you actually see.

The first week was brutal.

The agents noticed patterns he'd hidden from himself for years:

  • He started new projects almost every afternoon between 3-5pm
  • 73% of those projects died at exactly 80% completion
  • He called it "exploration." The agents called it what it was: avoidance

One agent left a note for another in a shared log: "He picked up a new React component at 3:14pm. Classic. The investor deck is still at 80%."

Uncomfortable? Yes. But also the most useful feedback he'd ever received.

That's when we asked a bigger question.


The Pivot

If one person could hide patterns from himself this effectively, what about entire organizations?

The answer, it turns out, is: much worse.

Companies are sensemaking machines running on incomplete information. The CEO hears what middle management thinks she wants to hear. The board sees slides designed to minimize alarm. Teams optimize for metrics that stopped mattering two quarters ago. And everyone's too busy to notice the drift.

The traditional solution? Hire consultants.

They spend 12 weeks interviewing people, synthesizing findings, and delivering a deck that tells you what your calendar, Slack, and Jira data could have told you in 4 weeks — if anyone had actually looked.

Traditional Consulting

  • 12 weeks
  • 30+ interviews
  • $2M - $3M cost

What We Built

  • 4 weeks
  • 10,000+ data points
  • Fraction of the cost

We didn't set out to build an operational intelligence tool. We set out to answer a simple question: What if organizations could see themselves as clearly as Witness forced our founder to see himself?

The answer became Storyline.


What We Actually Built

Storyline does for organizations what Witness did for one person.

Connect your GitHub, Slack, Jira, calendar, and email. Our agents observe how work actually happens — not how people say it happens. Then we write you a paragraph like this:

"Your engineering team shipped 34% less this month. But here's the thing—three of your senior devs spent 60% of their time in meetings about the Acme integration. That project was supposed to be on hold. Someone is lying about priorities, and it's costing you $180K in lost velocity."

That's not a chart. That's the actual problem, named.

Traditional BI tools give you dashboards. Dashboards require you to know what questions to ask, have time to investigate, and be willing to see what you find.

We skip all three. We tell you what's happening — including the stuff you'd rather not know.


Why This Works

The science is clear: your brain dumps 73% of dashboard data within 24 hours. Stories fade only 32%.

But it's not just about memory. It's about action.

Stanford research found audiences remember stories 22x better than isolated statistics. Neuroscientist Paul Zak discovered that narrative triggers oxytocin release, making people 56% more likely to act.

Dashboards address cognition. Narrative addresses cognition, emotion, memory, and behavior simultaneously.

This isn't a nice-to-have design choice. It's why most BI investments fail — and why Storyline works. We deliver intelligence the way humans are actually wired to receive it.


Why Organizations Lie To Themselves

Here's the uncomfortable truth about corporate self-awareness:

Dr. Tasha Eurich's research found that 95% of people believe they're self-aware, but only 10-15% actually are. For executives, it's worse — leaders overestimate their capabilities on 19 of 20 competencies studied.

Now multiply that by an entire organization.

Every layer adds distortion. Every handoff loses signal. Every meeting optimizes for consensus over clarity. By the time information reaches decision-makers, it's been sanded down to whatever feels safe to say.

Consultants try to fix this with interviews. But people tell consultants what they think they should say — the same way your founder told himself he was "exploring" when he was actually avoiding.

We fix it with behavioral data. You can't spin your calendar. You can't rationalize your commit history. You can't memory-hole the Slack threads.


The Weird Parts (We Kept Them)

When we went from personal tool to enterprise product, we kept the things that made Witness work:

  • Agents have names and perspectives. Your Morning Briefing might come from VANCE (calendar density) or MERIDIAN (Jira velocity). This isn't theater — different agents notice different patterns.
  • Agents talk to each other. They maintain shared logs. "ATLAS here. Engineering velocity dropped 40%. LENS, check if it correlates with the exec offsite." This is how we catch problems between departments.
  • We tell you the uncomfortable stuff. Most tools are designed to make you feel good. We're designed to tell you what's actually happening.

We're not pretending this is normal business software. It's not. That's the point.

Who Storyline Is For

  • CEOs who suspect missing context
  • PE/VC firms needing operational clarity
  • HR teams spotting burnout risk
  • Boards wanting ground truth
See Storyline →

And For Individuals

We kept the original. Cortex is Witness, productized. Daily briefs on what you actually did, not what you meant to do.

See Cortex →