Active Research

Our current experiments and research initiatives.

Learning Intelligence

Can ML predict when a learner will answer incorrectly?

Status:Active
Performance:AUC: 0.7067

Building ML models to predict learner performance, enabling early intervention and personalized support.

Grant Compliance AI

Can AI identify compliance requirements across funding docs?

Status:Production
Tests:82 automated

Developed NLP techniques to extract requirements and cross-check against proposals. This became GrantGuard.

Healthcare Data Quality

How can systems identify quality issues before decisions?

Status:Active
Validation:Real-time

Developing systems to automatically validate healthcare data and flag anomalies early.

Research approach

How we conduct research and move ideas into production.

Define the question

Start with a real problem, not technology looking for problems.

Gather data

Work with partners. Understanding data quality and limits is critical.

Experiment

Test approaches, evaluate results, document findings. Failure is part of it.

Validate

Test with real users in real environments. Lab success doesn't guarantee operational success.

Deploy or learn

If it works, productize it. If not, publish findings and explore next ideas.

Interested in collaborating?

Share data, co-research, or be an early adopter. Let's work together.

Get in touch ?