PRISM
/

EVERY PATIENT
SCREENED
BEFORE CLINIC.

PRISM screens every upcoming patient on tomorrow’s schedule and the week ahead against every active trial in your portfolio, then refreshes the results nightly as each visit approaches. Your coordinators don't read charts to rule patients out. They walk into clinic with a cited, prioritized worklist of candidates, every day.

TOP-3 TRIAL MATCH ACCURACY
95.5
88
95.5
%
ELIGIBILITY CRITERIA ACCURACY
92
88
92
%
REASONING ACCURACY
91
88
91
%
CITATION ACCURACY
94
88
94
%
Why this matters
/

Manual chart review is the only thing standing between most patients and the trials they’re eligible for.

A coordinator can read maybe a dozen charts a day. A clinic runs hundreds. The math has never worked, and it gets worse every year as protocols multiply and complexity grows. The result, every cancer center already knows: missed enrollments, slower studies, and patients who never got the chance.
How it works
/

Three matching modes. One worklist.

PRISM doesn’t just match. It re-evaluates eligibility every time the chart changes, and it surfaces the patient the moment they qualify.
1.1
Patient to Trial.
Every patient with an upcoming appointment is screened against every locally open trial. Matches arrive as an EHR in-basket message or in the PRISM worklist before clinic starts, with evidence pre-cited back to pathology, molecular, and note-level details.
1.2
Trial to Patient.
Spin up a new study or evaluate a feasibility opportunity, and PRISM identifies the active cohort eligible today. Feasibility in hours, not weeks.
1.3
Continuous Tracking.
Patients ineligible today don’t disappear. PRISM re-evaluates eligibility every time the chart updates, detecting progression, new biomarkers, or a change in line of therapy, and surfaces the patient the moment they qualify.
Outcomes
/

Production results, from a leading academic cancer center.

Patients screened monthly
7,000
88
7,000
Increase in trial accruals
30
88
30
%
Faster eligibility screening
70
88
70
%
Trials live in four weeks
100
88
100
%
principles
/

We built PRISM to be:

1.1
Proactive.
Matches can surface as in-basket or email notifications when the chart updates, not when a coordinator finally has time to look.
1.2
Comprehensive.
Federally funded, investigator-initiated, industry-sponsored. Your full portfolio, no exceptions.

1.3
Compatible
Native Integrations with Epic, OncoEMR and others. Approved on the Epic App Marketplace.Native in Epic, Cerner, OncoEMR. Approved on the Epic App Marketplace.

1.4
Scalable
No study-specific build effort. New protocols go live on day zero.

1.5
Accurate.
Peer-reviewed expert-level accuracy (Nature Digital Medicine, ASCO).

1.6
Equitable
At-scale prescreening reaches patients from underserved communities, meeting diversity enrollment targets without extra coordinator hours.

Trust
/
/
PRISM
·
A

Every match carries its evidence.

PRISM doesn’t just say “match.” It shows the pathology report line, the molecular result, the imaging note, the line of therapy, every piece of evidence behind the decision. Click any criterion to jump to the source. Research offices and IRBs can audit any output without leaving the workflow.
Learn more about PRISM
More by the numbers
/

Additional PRISM metrics.

Placeholders sitting on the live counter component. They render TBD until production values arrive, so the layout stays stable as numbers ship.
TRIALS LIVE ON PRISM AT LARGEST SITE: 300+
MONTHLY VOLUME SUPPORTED, ALL SITES 65,000 PATIENTS
AVERAGE GO-LIVE TIMELINE 14 wks
BEST GOOD to NOT-GOOD MATCH RATIO 2:3
Publications
/

Peer-reviewed, year after year.

PRISM’s science holds up under scrutiny. A selection of peer-reviewed work and accepted conference contributions.
2025

 2025 ASCO Quality Care Symposium

Equity through efficiency: Automating patient to trial matching using AI and OncoLLM. Demonstrated a >30% increase in treatment trial accruals.

May 28 2025

 ASCO 2025 Annual Meeting

Enterprise-level AI screening for oncology trials: Improving patient identification with OncoLLM. Demonstrated a 40% increase in trial matches.

28 October 2024

Nature Digital Medicine

PRISM: Patient Records Interpretation for Semantic clinical trial Matching system using large language models. Demonstrated a 95% accuracy in matching patients to trials.

07 June 2025

Central Surgical Association

Understanding Unrealized Trial Enrollments Following Patient-to-Trial Matching with OncoLLM in Surgical Oncology.

Get started
Triomics

SEE THE PLATFORM

LIVE IN ACTION

A 30-minute working session with our team. Use our synthetic-data sandbox, or sign a BAA to upload your own data.

Book a demo
Close

Book a demo today

A thirty-minute working session with our team. Bring a de-identified record or use ours.

Please fill out these details and we’ll get back to you.

By signing up, you accept our
privacy policy and terms of service.

Submission received.
Error submitting your form, please try again later.