The answers your firm already wrote, one question away
Hybrid retrieval over your documents, SOPs, precedents and transcripts. Every answer cited, every gap admitted.
Your custom strategy and systems, delivered in under a month. Your named people using them every week by day 30. Or we keep working for free until both are true.
Drawn from this system's own stages. A dashed ring marks a stage where a person acts.
01How it runs
Stage by stage.
When the corpus is silent, the brain says so. A logged gap costs you a question. A confident guess costs you a client.
- 01
Collection
Documents, SOPs, precedents and call transcripts gathered from the systems they live in today. Nothing gets retyped or moved by hand.
- 02
Indexing
Hybrid retrieval built over the whole corpus, so a question finds the right passage whether it matches the wording or only the meaning.
- 03
Answering
Every answer arrives with its source attached. You read the passage it came from, not a paraphrase you have to take on faith.
- 04
Refusal and routing
When the corpus does not hold the answer, the brain says so and routes the question to a person. The gap gets logged, not papered over.
A person acts here
- 05
Coverage measurement
The brain is tested against your team's questions. Gaps become an ingestion list instead of a surprise.
02What it does
Four jobs it takes on.
The firm answer desk
Anyone asks in plain language and gets the firm's own answer with the source attached.
Policy and precedent lookup
The passage that settles the question, found in seconds and cited back to the document.
Grounded refusal
When the corpus lacks the answer the brain says so and routes the question to a person.
Coverage audit
The corpus measured against your team's questions, gaps turned into a list.
What it replaces
The answer exists and nobody can find it.
The firm has answered this exact question before. It sits in a memo, a matter file, an old email thread. Redoing the work is faster than finding it.
Generic AI tools guess.
You tried a chatbot on the documents. It answered fluently, cited nothing, and was wrong in a way that mattered. Trust never recovered.
Your senior people are the search engine.
New hires interrupt the partner because asking beats searching. The most expensive person in the building fields the same question every week.
Knowledge walks out the door.
When someone leaves, the reasoning behind their decisions leaves with them. The documents stay. The map to them does not.
Nobody knows what the corpus covers.
You suspect there are gaps. Until coverage is measured against the questions your team asks, a gap only shows up when a client hits it.
03Proof
Where it runs.
Hoyes Michalos
LiveA cited answer from the firm's own material the day after kickoff.
Read the story1 day 6 hoursfrom repository created to the first cited answerSystem loggedby Advizr, 20 Aug 2026How this was measured
From the moment the repository existed to the first cited answer in the client's database. Both ends are timestamps the systems wrote themselves.
Measured at Hoyes Michalos knowledge brain
48 of 51answers grounded in a matching sourceSystem loggedby Advizr, 17 Jul 2026How this was measured
Every answer was graded by Advizr's harness: grounded, cited and drawn from the source that holds the answer. Of the three misses, one cited the wrong source and two declined a question they could have answered. The client's own grading sheet is blank, so this is our measurement, not their sign-off.
Measured at Hoyes Michalos knowledge brain, acceptance eval
8 of 8questions it had no source for, declinedSystem loggedby Advizr, 17 Jul 2026How this was measured
Eight questions had no source anywhere in the firm's material. The brain declined all eight: five with a plain refusal and three by saying what the material does not hold. An answer nobody can source is a liability, so a refusal is a pass.
Measured at Hoyes Michalos knowledge brain, acceptance eval
Nurse Next Door
LiveNurse Next Door asks years of franchise records in plain language, with the source shown
Read the story81,830records in one knowledge brainSystem loggedby Advizr, 20 Aug 2026How this was measured
The count of records in the brain's own table on the day it was read. The July reconciles read a little lower while the load was finishing. The records themselves are never shown.
Measured at Nurse Next Door knowledge brain
4 days 19 hoursfrom repository created to the first client sign-inSystem loggedby Advizr, 20 Aug 2026How this was measured
From the moment the repository existed to the first time someone at the client signed in. Both ends are timestamps the systems wrote themselves.
Measured at Nurse Next Door knowledge brain and portal
04What is included
In the build.
- Your corpus collected, cleaned and indexed on your isolated infrastructure
- Hybrid retrieval tuned on your real documents, not a demo set
- A citation on every answer, linked back to the source passage
- Grounded refusal wired in from day one
- A coverage report built from the questions your team really asks
- Ask from chat or the brain surface, wherever your team already works
- Training, monitoring and tuning after launch, inside the monthly partnership
05How it is built
Five layers under every system.
The same five layers sit under all 14 systems. Here is each one in this system's own terms.
Retrieval
Your documents, SOPs, precedents and transcripts are indexed for meaning and for exact wording, fused into one ranking. The passage behind every answer is retrieved before a word is generated, and retrieval quality is measured on its own.
- pgvector
- Full-text BM25
- Reciprocal rank fusion
Agents and orchestration
Ingestion, indexing and gap routing run as scheduled jobs, every run logged. When a question cannot be answered, the gap is filed and a person is notified rather than the question disappearing.
- agent-worker
- Scheduled ingestion
- Gap routing
Evaluation
The brain is graded with your people judging the answers. Coverage is measured against your team's questions and reported with the gaps listed.
- Eval graders
- Coverage sets
- quality-worker
Models
Frontier models write only from the passages retrieval returned. Embedding models handle meaning search. When retrieval comes back empty, no model fills the silence.
- Frontier models through one gateway
- Embedding models via the same gateway
- Grounded generation
Data boundary
Your corpus lives in your own database, one part of the isolated infrastructure each client gets. Access follows each person's role, and actions are recorded.
- Supabase row-level security
- Per-user scoping
- Training off without your consent
Named tools in this buildSupabase + pgvector, Eval graders, Slack
06How we build it
In this order.
- 01
Collect before you index.
The first weeks go to gathering the corpus from the systems it lives in and agreeing what belongs. If people know a document exists and the brain has not read it, they stop asking.
- 02
Prove retrieval before generation.
Retrieval is measured on its own against your team's questions before any answer is generated. If the right passage does not come back, no model downstream can save the answer.
- 03
Ship the refusal with the answers.
Grounded refusal is wired in from the first day, not added after an embarrassment. The brain earns trust by admitting what it does not know, and every refusal becomes a line on the ingestion list.
What we will not automate
Guessing. The brain will not fill a gap in your corpus with what a model remembers about your industry. An unanswered question goes to a person and onto the ingestion list, never into a confident paragraph.
07Questions
Straight answers.
What happens when the answer is not in our documents?
The brain says it does not know, shows the closest passages it did find, and routes the question to a person you name. The gap is logged so the missing document can be added. It will not guess.
Where does our data live?
In your own database, on the isolated infrastructure each client gets: its own repository, database and deployment, hosted by Advizr. Nothing you index trains a model unless your team turns that on. Sign-in is one shared service across clients.
How is this different from search?
Search returns documents and leaves the reading to you. The brain returns the answer with the passage it came from, so checking takes a glance instead of an afternoon. And when nothing in the corpus answers, it tells you that, which search never does.
How do we know the citations are real?
Every citation links to the source passage. You open the document and read the paragraph yourself. And when retrieval finds nothing, the brain refuses instead of inventing a source.
What goes into the corpus?
Whatever your firm relies on: documents, SOPs, precedents, meeting and call transcripts, and the notes buried in email and your CRM. We collect from the systems they live in today, and you approve the list before anything is indexed.
How do you measure whether it works?
Against your own questions, not a benchmark. We gather the questions your team really asks, run them through the brain, and grade the answers with your people. Coverage gaps become an ingestion list, and the same set runs again after every change.
Built alongside it
Team knowledge
Decisions and know-how captured as they happen, mapped and kept usable.
Outbound engine
Prospecting runs around the clock so your calendar fills itself.
Run your whole company on AI.
Tell us how your business runs. If we're not the right fit, we'll say so on the call.
Your custom strategy and systems, delivered in under a month. Your named people using them every week by day 30. Or we keep working for free until both are true.
The AI operating layer for enterprises.