Quillhand
Request early access
Self-hosted analytics · your data stays home

Ask your claims data anything.
Your PHI never leaves the building.

Quillhand is a self-hosted, HIPAA-friendly analytics platform for healthcare claims and revenue-cycle teams. Point it at your own database and ask questions in plain English — no analyst, no SQL, no data ever shipped to a vendor cloud.

Request early access See how it works

Query-in-place & read-only — Quillhand never copies or takes custody of your data.

Built for the way ops teams already work PostgreSQL MySQL SQL Server No-egress & HIPAA-friendly OIDC SSO
The daily bottleneck

You are sitting on a goldmine of claims data — and can't get an answer out of it.

Your billing system holds every claim, remit, denial, and payment. But the moment leadership asks a real question, the work grinds to a halt: someone has to pull it, someone has to know SQL, and nobody is allowed to send that data to an outside tool.

  • Reports take days, not minutes

    Every "quick question" becomes a ticket to IT or your software vendor — and you wait, then wait again for the follow-up.

  • No data team, no analysts

    Hiring a SQL analyst for a mid-market billing shop isn't realistic — so the questions pile up and the spreadsheets multiply.

  • Cloud tools mean shipping PHI

    Most modern analytics tools want to copy your data into their cloud. For a covered entity or business associate, that's a compliance headache you can't take on.

  • You can't ask a follow-up

    The static monthly report never answers the next question — "okay, but which payer, which provider, which month?"

“We have twelve years of claims history and I still have to email our vendor and wait a week to find out our denial rate by payer. Meanwhile the money sits there.”
— The reality for most mid-market RCM operations teams
From database to answer in four steps

An analytics team in a box — that sets itself up.

Quillhand does the work a data team would do: it learns your database, translates plain-English questions into safe SQL, and keeps the answers current. You don't write a query or move a single row.

  1. 1

    Point it at your database

    Deploy Quillhand inside your own network and connect it to your Postgres, MySQL, or SQL Server with a read-only account. Nothing is copied out.

  2. 2

    It learns your schema

    Quillhand profiles every table and column and builds a browsable semantic catalog — the business meaning behind claims, remits, and eligibility — so questions map to the right data.

  3. 3

    Ask in plain English

    “What's our clean-claim rate by provider this quarter?” Quillhand writes guarded, read-only SQL, runs it in place, and shows the answer and the query behind it.

  4. 4

    Pin it to a dashboard

    Save any answer to an auto-refreshing dashboard. Your denial trends, AR aging, and payer mix stay current — no rebuild, no monthly ticket.

The whole point of Quillhand

Your data never leaves your infrastructure. Not a copy, not a cache, not a single row.

Most analytics vendors ask you to trust them with your PHI. Quillhand is built so you never have to. It runs where your data already lives and queries it in place — so there's no third-party custody to worry about, and no business-associate exposure from shipping data out.

Self-hosted & no-egress

Quillhand deploys entirely inside your trust boundary — on-premise or in your private cloud. There is no Quillhand cloud that ever sees your rows. Your data stays exactly where it is.

Never takes custody of PHI

Because Quillhand queries in place and only ever reads schema metadata plus query results, it never copies or stores your protected health information — sidestepping the compliance burden of a data-shipping vendor.

Bring your own LLM

Use a BAA-covered managed model like Azure OpenAI or AWS Bedrock — or run fully air-gapped with a self-hosted open model via Ollama or vLLM. No prompt ever has to leave your network.

A guarded, read-only executor

Every generated query is parsed and AST-checked to prove it's read-only, then executed in a read-only session with hard timeouts and row caps. Quillhand structurally cannot insert, update, or delete anything.

Pair it with OIDC single sign-on and per-source access grants so every person only ever queries the data they're cleared to see — and every question is attributable. Compliance-conscious by construction, not by policy.

Built for healthcare operations

Made for the teams that run on claims data but never got a data team.

If you live in claims, billing, remits, and eligibility — and you've been waiting on someone else to answer your questions — Quillhand was designed for you.

TPAs

Third-party administrators

Track claim throughput, turnaround, and plan performance across every client without exporting a thing.

Billing / MSOs

Medical billing companies

See denials, AR aging, and clean-claim rates by client, payer, and provider — in seconds, not a support ticket.

MGAs

Managing general agents

Monitor loss ratios, program performance, and payer behavior across your book without a BI project.

Physician groups

Physician & specialty groups

Understand revenue per provider, no-show impact, and payer mix straight from your practice-management database.

And the engine is domain-agnostic — any mid-market operations team with a SQL database and no analysts can put Quillhand to work, in healthcare or beyond.

Everything the platform does

Meticulous by design — like a scribe who never misfiles a record.

Quillhand is opinionated about trust. Every feature is there to make the answer correct, current, and safe to act on.

Human-editable semantic catalog The moat

A living map of what your data actually means. Correct a definition once and every future answer gets it right — your institutional knowledge, captured.

Guarded read-only executor

AST-checked, read-only sessions with timeouts and row caps. The engine physically can't mutate your database — safe to hand to anyone.

Auto-refreshing dashboards

Pin any answer and it stays live. Denial trends, AR aging, and payer mix update on a schedule — no rebuilding, ever.

“Teach-it” inline curation

See an answer that's slightly off? Correct it in a click. Quillhand learns your terminology and gets sharper with use.

Per-source access & OIDC SSO

Grant access source-by-source and sign in through your identity provider. Everyone sees only what they should, and every query is attributable.

Day-one findings report

The moment Quillhand finishes profiling your schema, it hands you a starter report of notable patterns — value before you ask a single question.

Straight answers

Frequently asked questions

The questions compliance-conscious buyers ask us first. If yours isn't here, get in touch.

Does my data leave my network?
No. Quillhand is self-hosted and queries your database in place. It reads schema metadata to build a catalog and runs read-only queries against your own database — your rows, including PHI, never leave your infrastructure and are never copied to a Quillhand cloud.
Is Quillhand HIPAA compliant?
Quillhand is designed for HIPAA-regulated environments. Because it never takes custody of your protected health information, it avoids the business-associate exposure of shipping PHI to a third-party cloud. You deploy it inside your own trust boundary and, when you use a language model, you can choose a BAA-covered managed model or run fully air-gapped. Compliance ultimately depends on how you operate your environment, but the architecture is built to keep PHI in your hands.
Do I need a data team or to know SQL?
No. Quillhand is built for operations people who don't write SQL. You ask questions in plain English and get answers, tables, and charts. Quillhand writes the read-only SQL for you and shows its work, so you can trust the number and, if you're curious, see exactly how it was derived.
Which databases are supported?
Quillhand connects to PostgreSQL, MySQL, and Microsoft SQL Server — the databases that typically hold claims, billing, remit, and eligibility data. It profiles the schema you point it at and builds a browsable semantic catalog automatically.
Can it run air-gapped or on-premise?
Yes. Quillhand runs entirely on-premise or in your private cloud. For the strictest buyers it can run fully air-gapped with a self-hosted open model via Ollama or vLLM, so no request ever leaves your network.
How is it different from a cloud BI tool?
Traditional cloud BI copies your data into a vendor warehouse and still expects an analyst to model it and build dashboards. Quillhand stays inside your walls, learns your schema for you, and lets non-technical staff ask questions directly — no data pipeline, no analyst, and no egress.
Can Quillhand change or delete my data?
No. Every query is parsed and AST-checked to confirm it's read-only, then run in a read-only database session with strict timeouts and row caps. Quillhand cannot insert, update, delete, or alter anything in your database.
What LLMs can I use?
Bring your own. Quillhand works with any OpenAI-compatible model. Use a BAA-covered managed model such as Azure OpenAI or AWS Bedrock, or run a self-hosted open model through Ollama or vLLM for a fully air-gapped deployment.
Early access

Put an analyst inside your walls this quarter.

Join the early-access program and be among the first RCM teams to ask their own claims data anything — self-hosted, read-only, and safe by design. No data leaves your building to find out.

Prefer a walkthrough? Book a demo · We'll never share your address. No spam, ever.