Day One Proposal

Day One at the practice

Prepared for the owner, on referral.

What this is

A day inside your practice before you commit a salary to finding its problems.

You are already looking at a COO because the practice feels heavier than it should. The first thing a good COO would do is spend their first month walking your floor, finding where the practice leaks time and money. Day One is that walk, compressed and structured, done before the salary is committed instead of after the hire. If the day finds the leaks, you decide what to do about them with evidence in hand. Maybe that is still a COO. Maybe it is something cheaper and faster. Either way, the job description gets written from the diagnosis instead of the other way around.

We spend the day where the work happens. The reception desk and the phones. The inbox your website feeds. An attorney's afternoon, including what is waiting on the desk after a morning in court. The file room, the calendars, the systems that hold the matters. We watch the work as it happens, not as anyone describes it.

We come ready to listen and to think on our feet. No prepared deck. And do not tidy the floor for us. The mess is the data.

What you walk away with

A playbook, within two weeks of the day. Not a deck. Not a recommendation memo hiding in a PDF. A written read for operators.

The playbook answers three questions:

• Where AI fits in this practice, and where it does not.
• The highest-leverage moves we see, sequenced so you can act on them in order. A roadmap, not a list.
• What it would take to run the sequence: with your own team, with another firm, or with us.

The playbook is yours. Run it however makes sense.

What we need from you

• The owner's morning: an hour to start the day, and availability when questions surface.
• Time with the people who answer your phones and work your front office, with their normal day running around us.
• An attorney's afternoon. We want to sit with someone who was in court that morning.
• Access to the operational record: the phone logs, the website inquiry inbox, the case evaluation calendar, and the practice management system. Client matters stay privileged; we are watching the operation, not the cases.

The terms

A flat fee of [flat fee] for the day and the playbook. Travel and expenses billed at cost, on top.

No retainer. No commitment beyond the day itself. If we are the right fit for what comes next, we will already have been talking about what that looks like. If we are not, the playbook is still yours to run.

What happens next

After the playbook, you decide. Run it with your own team, hand it to another firm, or build it with us. If the work points to a build we are right for, we will scope it in a separate proposal once the playbook has shown what is worth building.

Day One Audit

The one-line finding

The practice feels heavier than it should, and the COO search is real evidence of that strain. The day says the weight is real, but it is not where an operations hire would look first. Your casework is disciplined, because statutes of limitation and litigation calendars allow nothing less: deadlines hit, filings move, cases get worked. The leak is upstream of the cases, at a front door nobody owns, where new injury inquiries wait [hours, sometimes days] for a first response in a market where the first helpful answer signs the client and the insurer's adjuster is already dialing. And the thing your clients actually pay eighty people for, the firm's accumulated judgment about what a case is worth and how to prove it, lives in a few senior heads and a thousand personal folders, so every question inside the firm routes to a person and interrupts the day of whoever knows. Fix the front door first. On contingency, the front door is the whole business: the firm earns nothing it does not first sign, and the marketing that generates every one of those inquiries is already paid for. Then write the firm down, so the knowledge serves the whole practice and the practice can see itself.

How we looked, and how we measured

One day inside the practice, structured. The morning at the front: the reception desk with the phones running, the website inquiry inbox, the case evaluation calendar. The afternoon with the people the work lands on: a senior attorney back from a morning hearing, the office administrator who carries the inbox, the staff who route what the phones catch. The measuring came from the firm's own record: the phone log, the inquiry inbox's timestamps, the evaluation calendar against the cases that actually signed. Where a figure below comes from that record, this audit says so. Where it comes from the legal industry's published benchmarks, secret-shopper studies of how firms answer their own phones and inboxes, it says that too.

What exists, and what the practice needs

The operating systems of a working injury firm are real and load-bearing. They start at the moment a case signs. Nothing structured exists before that moment.

The first row is why the firm has its reputation. The other rows are the project.

Stakeholder map

Each role holds a different piece of the problem. None is wrong. None alone can see the whole.

Note what is missing from the map: nobody owns intake. The receptionists answer what rings. The administrator clears what queues. The attorneys call back what reaches them. Intake is everyone's second job and no one's first, which is why its failures appear on nobody's report.

One inquiry, traced end to end

We traced the path a new case must travel, from first contact to a signed retainer agreement. Nine steps. The first five are where cases die.

1. The phone rings, or the form submits. During office hours, reception answers between transfers. Outside them, voicemail. [Roughly a third] of the week's inquiries arrive outside office hours, and injuries keep no office hours: a crash on a Friday night, a fall in a parking lot on a Sunday.
2. A message gets taken: a name, a number, a line about what happened. No screening. Reception is neither equipped nor positioned to ask what decides everything downstream: when and where it happened, the injuries and the treatment, the liability picture, the insurance on the other side, whether the limitation clock has nearly run, whether another firm already signed it.
3. The message routes by habit: to the attorney the caller asked for, or whoever comes to mind. No view of capacity. No match of case type to attorney.
4. The callback comes when court and evaluations allow, [that evening, often the next day], and then the phone tag starts: [two or three] attempts each way. The firm's own log shows inquiries that took [four days] to first contact, and inquiries that never connected at all.
5. The web form follows a slower road: a shared inbox, checked when the administrator has time, median first response [the next business day]. A form submitted Friday night waits [until Monday]. The injured caller does not wait with it: the adjuster calls within days, often first, and a quick settlement signed in a hospital hallway ends the case before the firm ever heard of it.
6. The conflict check runs manually before an evaluation books. Necessary, and another [day] in the queue.
7. The evaluation: an attorney's hour, free to the caller, not to the firm. The calendar shows [roughly a third] of evaluation hours spent on cases the firm would never sign: no liability, no recoverable damages, time-barred, already represented, outside the venues the firm works.
8. The retainer agreement goes out by hand and gets followed up inconsistently.
9. The case signs. From this step on, the practice is disciplined, and the calendars keep it that way.

The structural insight: a lost case leaves no trace. A signed case is managed loudly, in calendars and dockets and settlement reports. An inquiry that gave up after voicemail appears in no report anywhere. That is why this leak was invisible from the corner office, and why the felt diagnosis was "operations" rather than "the front door."

The friction, quantified

From the day's record, stated against the industry's published behavior.

The funnel, compounded

The stages multiply. Of [about 150] inquiries a week, [roughly a third] meet voicemail first and some share never returns. Of the messages taken, screening happens only at the evaluation itself, so attorney hours filter what a screen should have. The week we measured, the practice signed [about ten] new cases from [about 150] inquiries, in the neighborhood of [7 percent], which published intake benchmarks place at the slow end of an achievable range that runs past [the mid-teens] for firms that answer first and screen early. The gap between those two numbers, at the firm's average fee per resolved case, a contingency fee running well into five figures, compounds to [a seven-figure] annual difference. Every inquiry in that arithmetic was already bought and paid for by the firm's marketing. This is not growth that needs new spend. It is revenue the firm is currently declining by being slow.

What the day disproved

Leadership's working theory was an operations problem: the practice had outgrown its administration and was looking for an executive to run it. The day found the opposite shape. Downstream of a signed case, the practice runs with discipline; the limitation periods and the court calendars see to that, and the staff is genuinely good. What the practice outgrew is its front door, built for the firm it used to be, half the size. The second finding compounds it: as the firm doubled, its knowledge did not centralize, it dispersed. Every new attorney added a new set of personal folders, and the distance between "someone here knows this" and "the person asking can find it" grew with every hire. A COO would have found both leaks eventually, from inside a salary. The day found them from outside one.

Where AI fits, and where it does not

• Fits: the first response. An intake that answers in minutes, at any hour, in the caller's language, gathers what reception cannot ask, screens for merit, venue, limitation, and conflicts, and books the qualified onto the right attorney's calendar by case type and capacity. Speed is the product. In this market, the first helpful answer signs the client, and the other side is already calling.
• Fits: the firm's knowledge. The case valuation judgment, the procedure per court, the records and lien playbooks, the demand precedent that works, written down once, versioned, curated by the senior attorneys, and served to every attorney, clerk, and staff member who needs it. The interruption economy ends when the answer no longer lives only in a person.
• Fits: the picture of the practice. Caseload, progress, rulings, and status read from the systems the firm already runs, assembled into one view. Nobody does new data entry. The picture must be true without asking anyone to maintain it.
• Does not fit: legal advice. The intake gathers facts and explains process logistics. It never advises, never values a case, never predicts an outcome. That line is the practice's license, and it is enforced in the system's design, logged on every interaction, not promised in a policy.
• Does not fit: the evaluation, the strategy, the courtroom. The judgment clients pay for stays with the attorneys. The system's job is to buy that judgment back its hours.
• Does not fit: crisis. A caller still in treatment, in distress, or describing an unsafe situation is routed to a human immediately, at any hour, every time. No exceptions, no optimization.

Risks and constraints we observed

• The advice line is structural, not cosmetic. Everything the system says to a prospect must be information, never counsel and never a case valuation, with every word logged. The bar's rules on unauthorized practice and on solicitation set the boundaries of the build.
• Prospect information is sensitive before privilege even attaches, and in this practice it is medical. The screen must capture enough to qualify and no more than it must, and a conflicted prospect's story must not contaminate the firm.
• The receptionists are not the problem and the fix must not read as their replacement. The front door was unowned and unstructured, not lazy. The people who know the callers best belong inside the new system, not outside it.
• The knowledge effort fails if it becomes a wiki. Doctrine without an owner and a version is a rumor. The senior attorneys must curate or it must not ship.
• The clientele is in crisis. The caller is often days from a hospital bed, talking to insurers they should not be talking to. Every design choice gets measured against a person on the worst week of their life. Respect is a requirement, not a tone.

The signal we leave with

The first move is the front door, and it starts with cheap tests, not a build: shop the firm's own intake the way a stranger meets it, and run a two-week fast-response trial on one channel to prove what speed alone is worth. The evidence will set the build's shape and its targets. The second and third moves, the firm's knowledge written down and the practice made visible, follow on the same system. The plan, sized by impact, is the Playbook and Delivery Proposal.

Playbook and Delivery Proposal

The playbook and the delivery proposal are one document because they are one act. The playbook says where AI fits and sizes the moves in order. The delivery proposal scopes the build for the moves you choose to start with. The first earns the second. Nothing past the first move is committed until the first move proves the approach in your practice.

Part One: The Playbook

A written read for operators, not a deck. It answers three questions: where AI fits in this practice and where it does not, the highest-leverage moves in sequence, and what it takes to run them.

Where AI fits, and where it does not

It fits the work that is structured, repeated, and currently done by interruption: the first response to a new inquiry, the screening questions that decide everything downstream, the answers that live in senior heads, the status that lives in five systems. It does not fit the law. No machine in this build advises a client, values a case, predicts an outcome, or recommends a course; that line is the practice's license and it is enforced in design, logged on every interaction. The evidence is in the Day One Audit. The short version: the casework is disciplined and the front door is unowned, the firm doubled and its knowledge dispersed, and on contingency that combination is paid for every week in cases never signed and interrupted hours.

How to read the roadmap

The first three moves we diagnosed on the day, and we can size them against the firm's own record. The last two we saw the shape of and did not diagnose, and we say so rather than dress them up. Honesty about what is proven and what is a candidate is the difference between a roadmap and a sales sheet.

Each move is read across six dimensions: time, accuracy and quality, cost and recovered revenue, growth, employee experience, and risk. The first move earns the right to the next.

The roadmap at a glance

Move 1: The governed front door (start here)

An intake that answers in minutes, at any hour, on both channels. It greets the caller or the form in their own language, gathers what reception cannot ask between transfers, screens for merit, venue, the limitation clock, and the conflict basics, and books qualified cases onto the right attorney's calendar by case type and capacity. It explains process and logistics, never the law and never what a case is worth. A crisis disclosure routes to a human immediately, at any hour, every time. During the trust-building period, every inquiry the system would turn away goes to a person before it is turned away, because a wrongly declined client is the one error the firm cannot see in any report.

• Time: first response falls from [hours, sometimes days] to minutes, around the clock. The [third] of inquiries that arrive outside office hours stop meeting voicemail, and the firm starts beating the adjuster to its own callers.
• Accuracy and quality: screening happens before attorney time instead of during it. The [roughly a third] of evaluation hours now spent on cases the firm cannot sign becomes attorney capacity again, [15 to 20] hours a week of it.
• Cost and recovered revenue: the inquiries are already paid for; the marketing spend that generated them is sunk. Moving signing from the slow end of the published benchmark range toward its middle, on the firm's own funnel arithmetic, is worth [seven figures] a year at the firm's average fee per resolved case.
• Growth: the practice stops competing only during office hours. The firm that answers first signs the client; this makes the firm permanently first.
• Employee experience: reception stops apologizing for the queue. The attorneys' callbacks start with a screened, scheduled, conflict-checked evaluation instead of a cold slip. The people who know the callers best move inside the system as its reviewers and owners.
• Risk: the advice line and the crisis route are the two design absolutes, enforced and logged. Containment is structural: one channel first, every decline human-reviewed, every transcript auditable. Reversible at the cost of the build itself.

Move 2: The firm's knowledge, written down

The firm's accumulated judgment, captured as living doctrine: the case valuation patterns, the procedure per court, the records and lien playbooks, the demand precedent that works, the research already bought. Curated by the senior attorneys, versioned like law, attributed, and served to every attorney, clerk, and staff member who needs it, in seconds instead of interruptions. This is also the front door's brain: the screen's questions and routing rules draw from the same doctrine, so the firm's judgment reaches the first phone call without any attorney taking it.

• Time: the interruption economy ends. The question that cost a senior attorney's attention now costs a query. Research the firm already paid for stops being rebuilt at associate rates.
• Accuracy and quality: one current answer instead of [thirty] private versions. New attorneys onboard against doctrine, not against whoever has time.
• Cost and recovered revenue: research spend falls when the firm's own work is findable. Senior hours return to senior work.
• Growth: the firm's knowledge stops being a function of who is in the building. Scale stops diluting quality, which is the failure that quietly caps every professional practice.
• Employee experience: clerks and juniors get the senior attorneys' judgment without the senior attorneys' calendar. The seniors stop being the search engine.
• Risk: doctrine without an owner is a rumor. The senior attorneys curate, every entry carries a version and an author, and nothing uncurated ships. Contained by case type, one at a time.

Move 3: One picture of the practice

Caseload, progress, rulings, deadlines, and status, read from the systems the firm already runs and assembled into one live picture, visible to the attorneys, the clerks, and operations. Nobody enters new data; the picture is true because it reads what the work already produces. Workload balances on evidence instead of memory. The question "where are we on this" gets a screen instead of an interruption.

• Time: status stops costing attention. The owner reads the practice in minutes instead of assembling it from asks.
• Accuracy and quality: workload assignment by actual capacity instead of recall. Rulings, limitation dates, and deadlines visible to everyone they touch.
• Cost and recovered revenue: the senior attorneys' fragmented hours reassemble. Trial preparation stops being interrupted by questions a screen can answer.
• Growth: a practice that can see itself can grow without the owner becoming the bottleneck, which was the original strain behind the COO search.
• Employee experience: the interns and clerks see the whole practice for the first time, and operations stops working blind between the attorneys' calendars.
• Risk: lowest on the board. Reads from existing systems; changes nothing about how cases are worked. The risk is dashboard vanity, and the mitigation is ruthless: if a number does not drive a weekly decision, it does not appear.

The later moves: named, not yet diagnosed

• Move 4: Client status communication. Clients call because they cannot see their own case, and an injury case moves slowly enough to make the silence frightening. Proactive, attorney-approved status to the client, in plain language, on the rhythm the case actually moves. Not diagnosed; it earns a test only after move 3 makes status reliable enough to publish.
• Move 5: Drafting from firm precedent. First drafts of routine filings and demand packages assembled from the firm's own doctrine and the case's facts, always attorney-reviewed, never sent unread. Not diagnosed; it earns a test only after move 2's doctrine is live and trusted, because a draft from bad doctrine is worse than no draft.

Each is a contained bet with its own measurable result. None is committed now. They earn their turn only after the moves ahead of them prove out.

What it takes to run the moves

The discipline matters more than the technology.

• Draw the advice line in the architecture, not the policy. Information, never counsel, never a valuation, logged on every interaction. The line is configurable; crossing it is not.
• A human reviews every decline until the evidence says otherwise. Trust is granted per decision type, on catch rates, never by decree.
• Doctrine has owners. The senior attorneys curate the knowledge or it does not ship. Versioned, attributed, current.
• The picture reads; it never asks. No new data entry, ever. A picture that needs maintenance becomes fiction in a quarter.
• Crisis routes to humans, always. At any hour, without exception, faster than everything else in the system.

The plays that run each canon come from our reusable plays library. The ones selected for this engagement are instantiated in the Charter. ## Who runs it

This can run with your own team, with another firm, or with us. It needs a few clear accountabilities: the owner clearing the way and holding the why, a senior attorney who carries the doctrine's integrity, an intake owner who reviews what the screen declines and what the transcripts show, and a builder who runs the machinery. The judgment seats are the firm's. The machinery and the build craft are the pieces you would bring in.

The recommended first move and the 90-day frame

Start with the front door. The first 90 days: the cheap tests that prove what speed alone is worth, the screen built on one channel with every decline human-reviewed, the crisis route live from the first day, and the doctrine work begun with the senior attorneys so the screen's questions sharpen as the firm's knowledge centralizes. Prove the front door on one channel and the practice gains not just recovered cases but the pattern every later move follows: structured, contained, human-gated, measured.

Part Two: The Delivery Proposal

The proposal to build the playbook's first three moves: the governed front door, the firm's knowledge written down, and the picture of the practice. Scoped only after the playbook showed what is worth building.

What we understand

An eighty-person practice whose casework is disciplined and whose front door is unowned. A firm that doubled, and whose knowledge dispersed as it grew instead of compounding. Inquiries that wait [hours, sometimes days] in a market where the first helpful answer signs the client and the insurer is already calling. Attorneys interrupted into being the firm's search engine. An owner already looking for a COO because the weight was real, before the day showed where it actually sat.

What we will build

A practice operating system, not a tool. The front door: intake that answers in minutes at any hour, screens for merit and conflicts, routes by case type and capacity, with the advice line enforced in architecture and a human on every decline. The firm's brain: doctrine curated by the senior attorneys, versioned, served to the whole practice and to the front door's screen. The firm's eyes: one picture of caseload, progress, rulings, and status, read from the systems already in place. The measurement spine under all three: a weekly read of the front door and the practice, in minutes.

How we will work

Four phases, mapped to the WISER canons. Each phase is independently valuable, priced on its own, and earns the next. The engagement can stop at any phase boundary with value already in hand.

Phase 1: Interrogate

Cheap tests before any build. Prove the diagnosis and set the targets.

• Shop the firm's own front door as strangers, on both channels, at varied hours, and time everything.
• Run a fast-response trial on one channel for two weeks: a human answering web inquiries within the hour, business days only. Measure what speed alone does to evaluations booked.
• Baseline the funnel from the firm's own record: inquiries, connections, evaluations, signings, by channel and hour.
• Test a screening script at reception for a week, and learn what the front of house can and cannot carry between transfers.
• End of phase: the response gap measured, the value of speed proven on the firm's own numbers, the build's targets set as law.

Phase 2: Solve

Build the front door on one channel and earn trust on it.

• The intake screen live on web inquiries first: minutes to first response, around the clock, merit and conflict pre-screen, scheduling onto the right calendars.
• The advice line enforced in architecture, every interaction logged, the crisis route to a human live from day one.
• Every decline reviewed by a person before it is final. The intake owner role stood up from the firm's own front-of-house staff.
• End of phase: one channel answering in minutes with measured signing lift, a decline log proving the screen's judgment, and the firm deciding expansion on evidence.

Phase 3: Expand

The proven pattern, extended to the rest of the front door and the practice.

• Phones join the screen: after-hours first, then overflow, then first-line, with reception repositioned as the system's human gate instead of its bottleneck.
• The doctrine build with the senior attorneys: case type by case type, versioned and attributed, serving the practice and sharpening the screen.
• The picture of the practice assembled from the systems already in place: caseload, progress, rulings, status, deadlines. No new data entry.
• End of phase: the full front door answering in minutes, the first case types of doctrine live, the practice visible in one read.

Phase 4: Refine

Govern the system as it earns autonomy.

• Oversight tiers by decision risk: scheduling and routing graduate first, declines last, advice never.
• Weekly transcript and decline audits on a standing cadence, because a screen drifts the way a model drifts, quietly, while the aggregates look fine.
• Doctrine review rhythm with the senior attorneys: versioned updates, never quiet edits.
• End of phase: a governed front door and a documented practice, run by the firm's own people, with the judgment where it started, in attorney hands.

What we need from you

• The owner's mandate, visible to the whole practice: this is how the firm works now, not an experiment running in the corner.
• A senior attorney's hours for the doctrine, protected on the calendar, because the knowledge build fails politely and invisibly without them.
• The front-of-house staff inside the build: they know the callers, and the system needs what they know.
• Access to the operational record and the systems that hold it. Client matters stay privileged; the build touches the operation, not the representation.

Infrastructure

You provide the practice management system, the phone system, and the website the firm already runs; the build reads from and writes to what exists. We provide the intake machinery, the doctrine and picture systems, the integration, and the implementation. Prospect and client data stays within the firm's control, with the same confidentiality posture the practice already owes it.

Who is working on this

A senior practitioner who leads the engagement and owns the system design with your team, and the build itself. Your people fill the judgment seats: the doctrine, the declines, the law. Small team, close to the work.

Investment

Phased. Each phase is priced on its own so the engagement can stop at any phase boundary with value already delivered. The fee basis and amounts are held in the private client record. We did not fabricate figures for this anonymized record.

Charter

What a Charter is

Not a project plan. Not a requirements document that executes once and collects dust. A Charter is the memory that survives the chaos. Its value is the decision log: when someone asks a year later why the intake screen never answers a legal question, or why every declined inquiry passed through human hands for months, the answer is here, with the alternatives that were weighed and the evidence that settled it. The Architect keeps it current, same-day.

Metadata

Positions

The work was held together by clear accountabilities, not an org chart.

On a small team one person can hold several Positions. As the system proved reliable, a Position could be augmented by an AI agent inside documented constraints, with the human shifting from doing to directing and reviewing.

Objectives and constraints

The build specification: what the project set out to do and the lines it would not cross.

Scope

In scope: the practice's operation, from first contact to a visible, balanced caseload. The front door on both channels. The firm's knowledge as curated doctrine. The picture of the practice, read from the systems already in place. Out of scope throughout: the practice of law itself. No advice, no case valuation by machine, no filing sent without attorney authorship, no client representation touched. The build buys the attorneys' judgment back its hours; it never substitutes for it.

Objective and success criteria

Answer every inquiry in minutes at any hour, put attorney time only where attorney judgment is needed, and make the practice visible to itself.

Constraints

• The system never advises and never values a case. Information and logistics, never counsel, enforced in architecture and logged on every interaction.
• A crisis disclosure routes to a human immediately, at any hour, ahead of everything else the system does.
• Conflicts are pre-screened before the story gets told, and a conflicted prospect's information does not propagate.
• The systems the firm already runs stay. The build reads from and writes to them; it replaces nothing the practice depends on.
• Prospect and client data, medical details included, stays in the firm's control. Nothing trains on it, nothing leaves it.
• The front-of-house staff move inside the system, not out of the building.

Architecture and human-in-the-loop design

Three connected layers. The front door: an intake screen on web and phone that answers in minutes at any hour, gathers facts, screens merit, venue, limitation, and conflicts, and schedules qualified evaluations onto the right attorney's calendar by case type and capacity. The brain: the firm's doctrine, curated by the senior attorneys, versioned and attributed, serving both the practice's people and the screen's own questions. The eyes: one picture of caseload, progress, rulings, deadlines, and status, assembled from the practice management system and the systems around it, with no new data entry anywhere.

The human gates were absolute at launch: every declined inquiry reviewed by a person before it was final, every transcript auditable, the crisis route live from the first day. The intake owner, grown from the firm's own front of house, held the decline queue and the weekly transcript audit. The gates loosened only by evidence, tier by tier, under the Hierarchy of Agency, and the advice line never loosened at all.

Current state at the start

Carried from the Day One Audit. A disciplined casework operation behind an unowned front door: [about 150] inquiries a week met a receptionist between transfers or a web inbox checked when time allowed, first response ran [hours, sometimes days], and [roughly a third] of inquiries arrived outside office hours to voicemail while the insurers' adjusters worked the same callers in real time. Evaluation calendars carried [roughly a third] unsignable cases. The firm had doubled, and its knowledge had dispersed instead of centralizing: valuation judgment and procedure in personal folders, status answered by interrupting whoever knew. The ask that started the engagement was a COO. The day reframed it.

Decision log

The decisions that shaped the build, each with the alternatives weighed and the evidence that settled it. This is the part of the Charter that answers "why did we do it this way."

The decision and experiment record

The supporting narrative behind the log. The project ran the full WISER method. Witness had already found where AI fit; the project picked up at Interrogate and ran through Refine.

Interrogate

The first weeks were spent proving the diagnosis was worth a build, on the firm's own numbers. We shopped the firm's front door the way a stranger meets it: calls and web forms, business hours and not. The gap was as measured on the day, and worse after hours. Then the cheap test that settled the argument: for two weeks, one person answered web inquiries within the hour, business days only, nothing else changed. Evaluations booked from that channel moved sharply. Speed alone, no AI, no build, moved the number. That test set the build's target and its ceiling: if one human answering fast for two weeks did this, a screen answering in minutes around the clock had a floor under it.

Two paths died in the same weeks, and their deaths shaped the build. An answering service took messages without screening: the pile moved, the evaluations did not. A screening script at reception collapsed at exactly the busy hours when most inquiries arrive. Together they proved the requirement nobody had stated: the front door needed speed, screening, and constant availability at once, and no arrangement of people alone delivers all three.

Solve

The screen went live on one channel, web inquiries, with the lines drawn before the first conversation. It gathers facts, explains process and logistics, and never answers the legal question, even asked directly, even asked desperately. It never says what a case is worth. The refusal is designed, not awkward: it names what the evaluation is for and books one. A crisis disclosure routes to a human at any hour, ahead of everything else. Every inquiry the screen would decline went to the intake owner before it was declined, and the early queue earned its keep: the catches tightened the screen's rules week by week, and the rules' versions were logged like the doctrine they were becoming.

The first weeks produced the scene the practice still tells. The senior attorney who had spent years returning two-day-old message slips read the transcript of the screen at work at two in the morning: a caller rear-ended that evening, writing from a hospital waiting room, screened, conflict-checked, and booked onto the right calendar for the next afternoon, with a question about ongoing treatment asked gently and a plain-language warning not to discuss the crash with the other driver's insurer before the evaluation. Under the old front door, that caller was voicemail, and likely the adjuster's signature by Monday. Instead the case signed.

Expand

The phones joined by stages, after-hours first, because the after-hours third of the volume had the worst baseline and the cleanest measurement, then overflow, then first-line, with reception repositioned as the human gate the system escalates to rather than the bottleneck it overwhelmed. Each stage passed the same test before the next began: catch rates on the decline queue, transcript audits clean, booked evaluations holding.

The doctrine build ran alongside, case type by case type, the senior attorneys curating what the firm actually knows: the valuation patterns by injury and venue, the procedure per court, the records and lien playbooks, the demand precedent that works, the research already bought and buried in folders. The screen's questions sharpened as the doctrine grew, because they draw from it. And the picture of the practice assembled from the systems already in place: caseload, progress, rulings, deadlines, and status in one read, for the attorneys, the clerks, and operations. The owner stopped assembling the practice from asks. The clerks saw the whole practice for the first time.

Refine

Live was not the same as governed. Scheduling and routing graduated to light review first, on catch rates near zero. Declines graduated last, months later, and never fully: a sampled human review stands permanently, because the wrongly declined client remains the one error no aggregate would ever surface. The advice line never graduated at all; it is not a tier, it is a wall. The weekly audit found the drift the cadence exists for, recorded below, and the system's autonomy was earned back the same way it was earned first: by evidence.

Hierarchy of Agency

Three tiers of human oversight on what the screen decides, by risk. The intake owner holds the gates; the tier governs how much attention each decision type gets.

A decision type moves to a lighter tier only on evidence: a full review cycle in which the gate's catch rate on that type stays near zero. Scheduling and routing earned Tier 1 that way. Declines hold a permanent sampled review. The advice line is not in the hierarchy; nothing graduates to giving counsel. If a tier drifts, it falls back to heavier review.

Risk register

Drift and incident record

After a doctrine update tightened the firm's case acceptance criteria, the screen began declining a borderline category it should have routed to evaluation: low-impact collisions with delayed-onset injuries, which read as out of scope under the new language but which the firm in practice accepts and signs. Nothing in the aggregates moved; booked evaluations held, response times held. The weekly decline audit caught it inside [a week]: the intake owner flagged a decline-rate shift in one case category, the transcripts confirmed the pattern, and the doctrine language was corrected the same day, as a versioned change.

The lesson logged: a screen drifts the way a model drifts, quietly, in the blind spots of healthy aggregates. Audit the decisions like outputs, on a cadence, against the current rules, at the category level where drift actually lives.

Evolution history

How the oversight posture changed over time, and why.

Current status and what transfers

The system is live and the firm runs it. Full implementation was delivered in six months. A new inquiry, on either channel, at any hour, gets a first response in minutes: screened, conflict-checked, and booked onto the right attorney's calendar, with the crisis route standing in front of everything. The evaluation calendars arrive screened. The doctrine serves the practice and the screen both, curated by the senior attorneys on a standing rhythm. The picture of the practice is the operating read for the attorneys, the clerks, and operations: caseload, progress, rulings, status, and the front door's own numbers, weekly, in minutes.

What transferred is the judgment infrastructure, not just the tools. The decline gates and audits are run by the firm's own intake owner. The doctrine is the firm's, versioned and owned. The decision log is the firm's memory of why every line was drawn where it was. The owner who had been looking for a COO got the part of the operating help the practice actually lacked, the part that finds and fixes the leaks, built into the practice itself.

Outcomes

• First response to a new inquiry fell from [hours, sometimes days] to minutes, around the clock, on both channels.
• The after-hours third of the firm's inquiries stopped meeting voicemail, in the hours when the other side's adjusters used to get there first.
• Evaluation calendars arrive screened and conflict-checked; the [15 to 20] attorney hours a week that went to cases the firm cannot sign returned to casework.
• Signed cases from the same inquiry flow rose; the figure is held in the client record.
• Attorney hours on casework up across the practice.
• Legal research costs down: the firm's own work, findable, stopped being bought twice.
• The firm's knowledge centralized as versioned doctrine, curated by the senior attorneys, serving attorneys, clerks, and operations alike.
• The practice visible in one read: caseload, progress, rulings, and status, with workload balanced on evidence instead of memory.
• A receptionist became the firm's intake owner, holding the gates on the system that replaced the message pad.
• The firm runs cleaner today than it did when it was half the size.

Plays

The WISER plays this engagement ran, instantiated with the client's specifics. This is the index and what each produced. The high-value plays are held as standalone documents; the rest were applied inline in this Charter. | Canon | Play | What it produced | Source | |-------|------|------------------|--------| | Witness | Friction Mapping | The front-door friction map from the day's record | Standalone play | | Witness | User Flow Mapping | The nine-step inquiry trace, first contact to signed case | Inline in the Day One Audit | | Interrogate | Assumption Auditing | The register of beliefs tested, the COO diagnosis above all | Standalone play | | Interrogate | Experiment Selection, Logging | The secret shop, the fast-response trial, and the two killed paths | Standalone play | | Solve | Human-in-the-Loop Design | The gate architecture: declines, transcripts, crisis, the advice wall | Standalone play | | Solve | Quality Objective Setting | The response and screening targets set as law | Inline above | | Expand | Expansion Sequencing | Channel by channel, case type by case type, gate by gate | Inline above | | Refine | Drift Monitoring, Incident Response | The weekly audits and the decline-drift catch | Standalone play | | Refine | Hierarchy of Agency Design, Graduation | The oversight tiers and the evidence that moves a decision type | Inline above |

The front door is built, the doctrine is owned, and the practice can see itself. The candidates that wait their turn, client status communication and drafting from firm precedent, will be diagnosed the way everything here was: on the floor, against evidence, one contained bet at a time.