Live on a Gilead-funded oncology study at Emory University

Every value in a trial, perfect the moment it exists.

Nightingale checks and cleans trial data at the source, so it lands in your EDC analysis-ready, with an audit trail on every value.

A large Phase III trial, from first patient in to database lock.

Capture, normalize, clean, monitor
~90 days of cleanup
Database lock
First patient inData captureMonitoring & QC
~90 daysof cleanup before database lock
8.1 daysfrom patient visit to the EDC
~97,000queries on a Phase III study
~12 hrsweekly per data manager on queries

Published industry figures. See sources at the bottom of the page.

Working with

  • Emory University
  • Duke Clinical Research Institute
  • Georgia Tech
  • Grady Health System
  • Children's Healthcare of Atlanta
  • Thinq CRO
  • Georgia CTSA
  • Harvard Innovation Labs
  • Neuro20
  • Evoq Therapeutics

Trial data is still typed in by hand, then cleaned up months later.

58%
of ~97,000 queries per Phase III study change no data at all.Stokman et al., 2021
8.23% of values are wrong when a person re-keys a record by hand.Garza et al., 2026
$556K
$56K direct$500K revenue
lost per day a trial runs late.Smith, DiMasi and Getz, 2024

A late trial costs the sponsor money. It also costs the people in it, and the people waiting behind them, time they may not have. Finishing sooner is the point of all of this.

One layer from every source to your EDC.

Nightingale sits between every data source and the EDC your study already runs. Sites keep their existing workflow.

What goes inNightingaleWhat comes outStudy documentsProtocol, lab and pharmacymanuals, consent formsRegulationsFDA, EMA and ICHrequirementsTrial dataHospital records, labs,wearables, ePRO and paperCheckDocuments and rules checkedagainst each other, before enrollmentCleanEvery value standardized atthe source, as it arrivesProveAn audit trail on every value,from source to submissionRuns every day until database lockYour EDCMedidata, Veeva, Oracle or anyYour output formatExactly as specifiedAudit recordEvery value, every stepRegulatorSubmission-ready packageStorage stays with you.
1

Check the rulebook

Study documents checked before enrollment

2

Inventory

Existing records put to use

3

Find the gap

Exactly what still needs collecting

4

Run the data

Mapped, converted, checked, delivered

5

Watch it daily

Early warning on sites and endpoints

6

Prove every step

An audit trail on every value

For biopharma, CROs and site networks.

A database that locks weeks sooner.

  • Works with the EDC you run
  • Nothing new for your sites
  • Early warnings while there is time
  • Storage stays with you

Expected impact on a study.

TodayWith Nightingale
Queries
60% fewervs today
Manual data work
70% lessvs today
Manual re-keying
Removedvs 8.23% hand-entry errors
Visit to EDC
Minutesfor connected sources, vs 8.1 days
Deploy a second trial
2 weeksvs 4 weeks for the first

Query, labor and error figures are design targets from an illustrative large Phase III model, being validated on active studies. Deployment times are from live deployments.

What a late database lock costs.

$2.3Min direct cost and manual labor on one trial
Direct cost of delay
Manual data labor

Plus about in revenue not yet earned while the product waits to reach market.

Uses $55,716 direct cost per day of delay and ~12 hours a week per data manager at ~$70 an hour on queries and data cleanup. Revenue uses ~$500,000 per day. Illustrative; sources below.

Compliance and security.

GxP vendor auditCompleted
21 CFR Part 11Aligned
ICH E6(R3)Aligned
HIPAABAA available
SOC 2 Type IIIn progress

SOC 2 Type II audit underway. Full controls, policies and documentation are available in our Trust Center.

Not a databaseNot a CRONot a regulated deviceStorage stays with you

Common questions.

Do we have to replace our EDC?

No. Nightingale is not a database. It delivers into whichever EDC the study already runs, including Medidata Rave, Veeva Vault, Oracle Clinical, REDCap and Castor, in exactly the format you specify.

Do our sites have to change anything?

No. Sites are never asked to do anything new and never become a customer. Their re-keying workload goes down.

Is AI making decisions about our data?

No. AI reads documents, notes and scans and proposes a value. Every value is then checked against inspectable rules, and the pipeline that maps, converts and delivers data contains no LLM. The same input always gives the same output.

How do we know it works on our studies?

Start with one. We baseline a comparable trial you have run, deploy Nightingale alongside on one study, and compare query volume, first-pass resolution and lock date before you commit a portfolio.

How is it priced?

On the inputs you already know: patients, data sources, phase and duration, so every quote is reproducible. For CRO partners, we quote a partner rate and you set the price.

Run your next study on Nightingale.

Twenty minutes is enough to tell whether the fit and timing are right.

Pick a time for an intro call.

Open scheduling page
Sources for figures on this page
  1. Stokman P. et al. (2021), Journal of the Society for Clinical Data Management: 20 Phase III studies, ~96,980 queries per study; 42% resulted in a data change.
  2. Garza M. et al. (2026): manual transcription error rate 8.23%; integrated capture showed no detected transcription errors.
  3. Smith Z., DiMasi J., Getz K. (2024), Therapeutic Innovation & Regulatory Science: direct cost of trial delay $55,716 per day; opportunity cost on the order of $500,000 per day.
  4. Tufts CSDD, eClinical Landscape Study: median 8.1 days from patient visit to EDC entry.
  5. Salary.com, US Clinical Data Manager compensation (September 2026): ~$70 an hour.
  6. Vendor-sponsored industry survey, n=88: ~12 hours a week per data manager per study on query and data cleanup work. Indicative.