Skip to content
Raj Bhensdadiya

Docubark

Third-party risk management, minus the analyst reading every vendor questionnaire by hand.

Role
Full-Stack Engineer
Timeline
12 months
Live site
Visit the live site
Vendor docsQuestionnairesExtractionNLP, MLRisk indicatorsAssessmentsame day
How it works: vendor documents and questionnaires are read automatically, risk indicators are pulled out with NLP and machine learning, and the assessment is written the same day instead of over weeks.
vendor review, down from weeks
Same day

Overview

An AI platform for third-party risk management (TPRM). It reads vendor documents and questionnaires so an organization can score its vendors and keep watching them.

The problem

Third-party risk assessment was manual. Reading one vendor's documents and questionnaires took weeks, and two analysts rarely landed on the same answer.

What I built

Built a TPRM platform that reads vendor documents on its own, pulls out the risk indicators, and writes the assessment, using NLP and machine learning.

Results

A review that used to run for weeks now finishes the same day, and nobody reads every page hunting for the risk indicators.

Screenshot of the Docubark product site, showing a vendor risk dashboard with inherent, security and residual risk ratings
The live product site.

Stack

  • React
  • FastAPI
  • JavaScript
  • Python
  • PostgreSQL
  • AWS

Built by Raj Bhensdadiya, an AI & full-stack engineer in Montreal. The full work history and the other projects have more, and email is the fastest way to reach me.

More work