Join the Peel pilot: we give you one tool of your choice and hook up your lab — instruments, LIMS, ELN, datasets — and we automate the integration that normally costs a fortune. Everything runs grounded and source-pinned: answers carry receipts, and when the floor can't verify, it says so instead of guessing.
Every application is read by a person on the team — no waitlist bots.
Runs on your hardware; access to your systems expands only after evidence.
Answers carry receipts. When the floor can't verify, it says so — it doesn't guess.
Peel is a research prototype from Perslis Research. Pilot access is granted in stages after a scoped onboarding review.
Hooking a lab's instruments and systems into anything modern is normally a bespoke engineering project that runs into six or seven figures — per tool. We automate that integration, so the pilot is a monthly plan instead of a capital expense.
Most lab-critical software predates modern APIs. Integrating it by hand is the expensive part — and it's exactly what we automate away.
| Integration path | Typical cost | Why |
|---|---|---|
| Legacy tool, no endpoint API | $75k – $1,000,000 / tool | Custom drivers, reverse-engineered file formats, months of bespoke engineering |
| Tool with API endpoints | Dramatically lower | A documented, callable surface — far less to reverse-engineer |
| Older technology | Cost climbs the further back you go | Fewer docs, deprecated formats, no maintainers — more to rebuild by hand |
| Typical lab "hookup fee" | $200,000+ | Billed up front, before you get any value |
| Perslis pilot | $15k/mo → $3k/mo | We automate the integration — no upfront build, no bespoke team, no six-figure invoice |
What the industry bills $200,000 to set up — and $75k to a million per legacy tool — we automate, and roll into a monthly you can actually plan around.
Tools that already expose API endpoints hook up dramatically faster and cheaper. The steep costs above are for legacy instruments and systems with no modern interface — the ones every lab has, and the ones we specialize in reaching.
The pilot is a working partnership, not a demo queue. We prioritize labs that will connect real systems, define a real outcome, and tell us where the runtime holds and where it breaks.
We are not looking for spectators. We are looking for a lab with a question worth answering correctly.
Before access is enabled we confirm who you are, that you're authorized to use the systems in scope, and how the pilot is contained. We never ask for passwords, private keys, patient identifiers, or confidential source material in this form.
Professional identity, institutional affiliation, and a credible public or referred footprint.
Confirm you own or are permitted to connect every system and dataset in scope.
Classify sensitivity — public, internal, confidential, or PHI/regulated — and the blast radius.
Allowlisted paths, tools, retention, and the sandbox boundary agreed before activation.
This is a private product-security and access review — not a government clearance, and Perslis does not issue one. Access can be paused or revoked if scope, authorization, or risk materially changes.
Specificity wins. Tell us what you study, what's currently slow, unreliable, or impossible, who owns the systems, and how we'd both know the pilot worked.
Four minutes on the live console: how it answers without weights, how it learns from its own failures, and what it does when a database is down.
Research prototype. Pilot program open now.