Our solution · The core engine

Automated medical coding

Codes the full encounter from the clinical note: CPT, E/M, ICD-10, HCC and DRG.

Clinical note → Code + confidence score

How it works

From note to codes

01

Clinical note in

The engine reads the full encounter documentation.

02

Code prediction

Each predicted code is returned with a confidence score.

03

Automate or route to a coder

During local model validation, we pinpoint the confidence score that achieves 95% accuracy. Predictions above that threshold are automated, predictions below route to your coding team.

Code sets returned

CPTE/MICD-10HCCDRG

Each code is returned with a confidence score from a model trained and run locally on your own data.

What your coders see

A code prediction

Urology · UreteroscopyAutomated · above threshold
Note textPre-op diagnosis: Left ureteral calculus, 8mm, proximal ureter. Procedure: Ureteroscopy with holmium laser lithotripsy and stent placement. Findings: Calculus identified and fragmented with laser; basket extraction of fragments; mild mucosal edema noted; 6Fr double-J stent placed.
Predicted CPTConfidence
52356
92.0
52356 ; 74420
7.9
52281
0.0
Site threshold for 95% accuracy, set during validation
Illustrative example

Let’s chat

See your own validation results

We will validate a locally-tuned model from your data. 80% of data is used for tuning, 20% set aside as a clean validation test set.