Book a discovery call
Seven stations, one continuous line. We learn how your work actually runs, find the real constraint, test before we commit, then build the fix from proven techniques and new technology. And we come back: the line loops, because your business doesn't hold still. The method came from 20+ years in work where every step had to be documented, tested and defended.
Our name is the process. We discover before we recommend, prove the path before we commit, then assemble the solution from the right parts. Nothing gets built until we understand what's actually broken, and nothing is finished the day it ships.
The problem as the business experiences it, before anyone has framed it.
The most expensive mistakes come from solving the wrong problem well.
The complaint is a symptom. We're after the constraint underneath it.
The cheapest way to kill a bad approach is to try it at low stakes.
Assembled for you, not pulled off a shelf. New technology where it earns its place.
A solution nobody can run without us isn't finished.
Working on handover day is a low bar. The question is whether it still earns its place a year in. Every other engagement ends at delivery. Ours returns on a set cadence to inspect what we built against how the business now runs, and revise it where it has drifted. That's the loop: inspection feeds back into Discovery, and the line runs again on what has changed.
Delivery is a moment. Usefulness is a moving target. Volumes grow, rules change, people join and leave, and a process that fit perfectly at handover starts to chafe. Adaptation is the standing commitment to change what we built so it keeps fitting — the workflow rewritten, the automation retuned, the governance updated. Every change documented and testable the same way the original was, so what you own stays defensible.
The process held together by spreadsheets.
It works because one person knows how it works. When they're out, it stops. When they leave, it breaks.
Data nobody trusts.
Three systems, three answers, and a monthly meeting spent arguing about which number is right instead of what to do about it.
Software that makes you work its way.
You bought the tool, then rebuilt your process around its limitations. Now the workarounds are the job.
Work that takes three days and should take two hours.
Usually it's not the work. It's the six handoffs, the re-keying, and the waiting in between.
A pile of documents and no way through it.
Contracts, claims, records, intake forms. We spent two decades making sense of exactly this, at a scale most businesses never see.
"We should be using AI for this."
Maybe. We'll tell you where it actually helps, where it doesn't, and what has to be true before you trust it with real work.
We map how work actually moves through your company, find where it stalls, then redesign and implement the flow that replaces it.
Discovery-ledPurpose-built applications and integrations for the problems off-the-shelf tools can't solve, shaped around how your team already operates.
AssemblyWhere AI earns its place in document-heavy, judgment-heavy work, with the validation and audit trail to prove the output holds up.
AssemblyPolicies, retention, security and structure for the data your business runs on, so it's an asset rather than a liability.
Every stationCertified preservation and collection when it matters: disputes, departures, investigations, audits. Not only for law firms.
SpecialistFractional technology leadership, vendor selection, testifying expertise, and strategic guidance for companies without a full-time CTO.
SpecialistWhere the method was built. Matter management, review platforms, production workflows and defensible process for firms and legal departments. See the detail
SpecialistMost process advice is written by people who never had to defend it. Ours came out of two decades in work where you did — enormous, messy datasets, deadlines that didn't move, and a standing requirement to explain every step you took and why. "Trust me" was never an acceptable answer.
That leaves a mark on how you work. You stop believing the org chart and start watching what people actually do. You measure before you recommend. You document as you build, because someone will ask how you got there and "I remember" isn't a method.
Learn the terrain before you touch anything. Assemble the fix from the right parts. Document it so it holds up. Then come back and keep it fitting.
Those habits turn out to have very little to do with the industry that produced them. The slow workflow, the data nobody trusts, the process held together by spreadsheets and one person's memory — that shape of problem is everywhere. We've been solving it for twenty years. The sector changed. The method didn't.
The founders
Two people, nearly twenty years
The two of us met almost twenty years ago, working on the kind of problems where being approximately right is the same as being wrong. We were not looking for a business partner. We were solving the same problem from different directions and kept arriving at the same answer.
That kept happening. Across different companies, different roles and different decades, we found we could hand each other a half-formed idea and get back something better built than what we started with. Neither of us has ever been interested in technology as the point. It's what the technology lets a group of people stop worrying about that makes it worth building.
Two decades of that produced a short list of things we won't bend on: learn the terrain before you touch it, prove what you claim, and leave people able to run what you built without you. Assembly Discovery is those habits with a name on the door.
Type one sentence about something slow, manual or untrustworthy in your business. The terminal runs a miniature discovery pass and drafts a findings memo.
We don't prescribe before we understand. We'll tell you if what you asked for isn't what you need.
Every business is built differently. New technology where it earns its place, trusted techniques where they're proven, never change for its own sake.
We came up where any step could be pulled apart and questioned. We document, test, and stand behind what we build.
We'd rather show you a finding than sell you a vision. Claims get evidence. Results get measured.
Two decades in the highest-stakes corner of data work taught us that trust is won on the details.
Every industry builds its own jargon, and technology builds the worst of it. We explain things in words you can repeat to your own team.
The first conversation is free and genuinely useful. Tell us what's slow, what's manual, or what you've stopped trusting, and we'll tell you whether it's something we can fix.
Step one / free
Thirty minutes. We ask about the work, you tell us where it hurts. No deck, no pitch.
Book itStep two / scoped engagement
A fixed-scope, fixed-fee engagement that ends in a written findings document you own, whether or not you build with us.
Ask about scope