APPRENTICE

Your scientific intuition, made computable.

CORTEX captures scientific intuition and embeds it in Apprentice: a reasoning partner shaped by your judgment, adapted to your workflows, and aware of your lab's full context. Unlike generic tools trained only on public data, Apprentice keeps longitudinal memory relevant to your science and gets more useful with every interaction.

The best scientists know more than they can articulate. They develop instincts about which results to trust, which methods to rely on, and which findings warrant conviction versus caution. This intuition is the accumulation of years of experimental judgment, and it shapes every decision a lab makes. No existing tool captures it. Apprentice does, and gives it back to you as a collaborator that reasons about your science the way you do.
We model this intuition through four constructs.
Knowledge graph
Construct 01

The Research Record

Everything you've done, stated, and published.
Structured and connected.

This is where the profile begins. CORTEX ingests a researcher's publications, methods sections, experimental logs, and datasets, and builds a structured map of their scientific footprint: the research programs they've pursued, the hypotheses they've tested, the techniques they've used, the collaborators they've worked with. Think of it as a complete, searchable history of a researcher's scientific life, a connected web of decisions and discoveries.

Waveform signal A solid waveform with a gradient highlight sweeping left to right and holding at the end.
Construct 02

The Method Signature

Not just which techniques you use. Which ones you trust.

Every researcher has methods they rely on and methods they treat with caution. The Method Signature captures this distinction by analyzing how each technique actually appears in a researcher's work: Is it their primary evidence or a supplementary check? Do they describe it with confidence or hedge the results? Have they refined the protocol over time or adopted it recently? The result is a calibrated profile of methodological commitment: what this researcher actually bets their conclusions on.

Construct 03

The Epistemic Fingerprint

How you reason. Your personal rules of evidence, made explicit.

Two researchers can look at the same data and reach different conclusions. Not because one is wrong, but because they apply different standards of proof. One researcher requires three independent validations before stating a mechanism. Another moves from correlation to hypothesis faster but designs more falsification experiments. The Epistemic Fingerprint captures these researcher-specific reasoning patterns: what counts as sufficient evidence, how they move from observation to claim, and what inferential rules they apply that no one else in their field necessarily shares.

Construct 04

The Belief Map

What you currently hold to be true, and how confident you are in each piece.

The Belief Map is a living model of a researcher's current scientific worldview. Every active hypothesis, every established conclusion, every open question, weighted by how settled each one is. It updates automatically as new results come in, and it's shaped by the Method Signature: a result from a technique the PI deeply trusts moves the needle more than one from a method they've always hedged on. When new data arrives, CORTEX doesn't just store it. It shows you how it should shift what you believe.

Core Capabilities

01
Natural-language traversal
Query the lab's full data landscape in plain language.
02
Context-aware literature search
Personalized to your domain, not the whole field.
03
Pipeline automation
Existing R and Python scripts: four tools become one prompt.
043
Provenance & reproducibility
Every experimental decision tracked end to end.
05
Knowledge preservation
When a researcher moves on, their judgment stays.
06
Faster onboarding
Months of osmotic context, compressed to weeks.