AI systems for science, healthcare, and human progress.

An applied AI lab building intelligent systems that institutions can trust and use every day.

Research programmes

What the lab is working on.

Each programme starts from a real institutional problem. The ones that hold up under evaluation become products.

Clinical communication simulation

DialogRxby NMI Labs AI

AI virtual-patient training for pharmacy education.

Voice-driven virtual patients that stay in character, listen before they respond, and are assessed consistently against a rubric. It's the first programme from the lab to reach everyday classroom use.

Real-time voiceRubric evaluationIndian languages
Explore DialogRx
DialogRx virtual patient session
Document intelligence

Glyph

Healthcare documents, read into structured data.

Reads handwritten prescriptions, lab forms, reports and bills, then shows confidence on every field and links each one back to its place on the page, so a person can verify quickly.

HandwritingPrescriptions and lab formsHuman verification
Explore Glyph
Glyph · Prescription · sample data1 line to review
Patient
Sample Patient
Age · Sex
42 · F
Doctor
Dr. Sample Name
Date
12 Oct 2026

3 fields hidden by your settings. Never sent to the model.

MedicineFrequencyDaysConfidence
Metformin500 mg · Tablet1-0-13096%
Amlodipine5 mg · Tablet0-0-13092%
Azithrol500 mg · TabletSuggestion: Azithromycin1-0-03Review · 58%
How NMI systems work

From input to informed action.

Every system we build follows the same path. We connect the right modalities, knowledge and actions, then keep experts in the loop where their judgement matters most.

  1. 01 · Ingest

    Understand context

    Documents, voice, data and conversations become usable context.

    Document intelligenceMultimodalVoice AI
  2. 02 · Reason

    Connect knowledge

    Grounded models retrieve, synthesise and explain what is relevant.

    Generative AIKnowledge systemsMachine learning
  3. 03 · Orchestrate

    Coordinate work

    Agents and automation move structured workflows forward.

    Agentic systemsIntelligent automation
  4. 04 · Assure

    Keep people in control

    Review, provenance, privacy and accountability are designed in.

    EvaluationHuman reviewProvenance
From the blog

Notes from the lab.

Perspectives on applied AI, evidence and the conditions that make technology useful in high-consequence work.

Healthcare AI

The Evidence Layer: agentic AI in healthcare

Why agentic AI in healthcare needs an evidence layer that clinicians can inspect, question and trust.

Read the article
Build what matters

Let’s make scientific intelligence operational.

Tell us about the workflow, curriculum or research problem you’re working on. We’ll explore it with you on a 30-minute discovery call.