We Bring AI with Innovation to Application and Products.

We're a small research-driven team from the University at Buffalo exploring how modern AI can responsibly support clinicians. Our work sits at the intersection of machine learning, human factors, and healthcare workflows. We care about rigor, safety, and clarity—shipping ideas only when they're understandable and verifiable.

Research-drivenArtifical IntelligenceMachine LearningSecurity-Privacy

Mission in MIDAS

MIDAS (Multimodel Intelligent Dermatology Analysis System) was built on an Idea, where blending AI's neurals to the Deep Neural thinking on Analysis Aspect trained to perfection

At MIDAS, our mission is to empower healthcare professionals with trustworthy, multimodal AI tools that enhance clinical decision-making in dermatology. By combining dermoscopic imaging with patient metadata, we deliver insights that are not only accurate but also contextually informed — helping clinicians detect and classify skin conditions with confidence and clarity.

MIDAS decision support

Our Team

People who turn research into reliable software

AS

Ajitkumar Senthil Kumar

ML Engineer • Data Science @ UB

Builds end-to-end MLOps: data pipelines, training, evaluation, and GPU inference services.

MLOpsPyTorchRAGFastAPI
MM

Manoj Maheshwar Jagadeesan

Research Engineer • Artificial Intelligence @ UB

Focuses on multimodal modeling, attention mechanisms, and clinical evaluation workflows.

MultimodalAttentionCVEvaluation

Engineering Stack

Modern technologies powering our AI solutions

Live Technologies
PyTorchtimmOptunaFastAPIPostgreSQLDockerMLflowWeights & BiasesVercel/RenderTailwindshadcn/uiPyTorchtimmOptunaFastAPIPostgreSQLDockerMLflowWeights & BiasesVercel/RenderTailwindshadcn/uiPyTorchtimmOptunaFastAPIPostgreSQLDockerMLflowWeights & BiasesVercel/RenderTailwindshadcn/ui
11
Technologies
100%
Open Source
24/7
Monitoring

Governance & Quality

Our commitment to excellence through structured processes and ethical AI practices

🔄

Dataset Versioning

Complete Audit Trails

Every dataset change is tracked with comprehensive versioning, experiment lineage, and full audit trails through MLflow integration.

Key Features

Version Control
Audit Trails
MLflow Integration
Data Lineage
Status
Active
⚖️

Bias Monitoring

Fairness & Equality

Continuous monitoring with calibration curves and comprehensive per-subgroup metrics to ensure equitable AI outcomes across all demographics.

Key Features

Calibration Curves
Fairness Metrics
Subgroup Analysis
Bias Detection
Status
Active
🔒

Privacy Protection

Security First

PHI-safe workflows with robust encryption, role-based access controls, and full compliance with healthcare regulations like HIPAA.

Key Features

HIPAA Compliance
Data Encryption
Access Control
Privacy by Design
Status
Active

Collaborate with us

Interested in research partnerships, Building-AI applications, Talk-Tech, clinical pilots, or security reviews? We’d love to talk.