01
Anatomy Lab
The Human Atlas: a real 3D human anatomy explorer with male and female bodies, a rebuilt zoom experience and translation into any language, under NeuroParsec Lab branding.
Home / NeuroParsec
NeuroParsec
AI for clinical care
and medical education.
NeuroParsec is my HealthTech company: built from first principles to change how care is delivered, starting with the tools clinicians and students actually use.
The mission
Technology that heals.
Healthcare runs on knowledge: and knowledge should be instant, visual and trustworthy. NeuroParsec builds AI systems that bring clinical understanding to the point of care and the classroom.
No gimmicks. Real tools, engineered carefully, tested in the real world.
What we're building
In the lab.
02
Clinical AI
AI assistants that help clinicians reason faster: decision support, not decision replacement.
03
Med Education
Learning experiences that make complex medicine intuitive: built with educators, for students.
The builds
From the lab to working products.
01
NeuroParsec Lab, in active development
Human Atlas
An interactive 3D anatomy explorer that turns the whole body into a hands-on classroom: rotate, zoom, take it apart, search and isolate a single structure. My layer: integrating the female body, rebuilding the zoom experience, adding translation into any language, and shaping it as a NeuroParsec Lab study tool. The planned adaptation for cancer education is the next stage, not a finished feature.
02
Working demo
Clinic Knowledge Assistant
A clinic's own documents turned into an assistant that answers staff and patient questions with sources, and rewrites doctor instructions into simple Bangla a patient can actually follow. The demo edition runs on a fictional clinic, Rahat Dental & Care, Dhaka, so the whole flow can be shown in one meeting.
03
In development, spec stage
Clinic Suite
The next step up for clinics: appointments, intake and follow-up on a proper health-data foundation. The specification is written and the architecture is chosen; the demo is not running yet, and this page will say so until it is.
Credits, stated plainly. Human Atlas: built on the Human Atlas viewer by ashemag (MIT); female anatomy from the Human Reference Atlas (HRA) and HuBMAP dataset by Kristen Browne and Heidi Schlehlein (CC BY 4.0). Clinic Knowledge Assistant: built on AnythingLLM by Mintplex Labs (MIT). Clinic Suite: built on Medplum (Apache-2.0). The product layers described above are the work of NeuroParsec and Simanto Mitro.
Clinic Knowledge Assistant
Patient Education Mode.
Clinics answer the same questions all day: prices, appointment rules, aftercare steps. The answers sit in paper files and in the doctor's head, so patients leave confused and aftercare is forgotten by the time they get home.
The assistant loads the clinic's own service list, prices, policies and aftercare sheets once, then answers from those documents, with the source shown. In Patient Education Mode, doctor instructions are rewritten into simple Bangla, in short sentences and everyday words, with every number, time and safety warning kept unchanged.
- For staff: instant answers from the clinic's own documents
- For patients: instructions they can actually follow
- For the clinic: its documents stay its own, and the setup is one product: content loading, configuration and handover
Also from this desk
Watson, my own AI assistant.
Watson is the personal AI assistant I designed and built to run my own working life: voice-first, no hands, holding memory across all my projects and working through my real browser and tools while I direct it.
Two lines run in parallel: the shipped v1.11 series, running locally on my own machine as a private service, and a next-generation engine in development. Watson operates private accounts, so it stays private: no public demo, by design.
Voice first.
Shipped v1.11 series in daily use. Next-generation engine in development.
How I work
First principles, real impact.
01
Understand deeply
Every build starts with the real clinical workflow: not the demo version of it.
02
Prototype fast
Working software early, tested with real users, iterated in the open.
03
Ship carefully
In healthcare, quality isn't a feature: it's the whole product.
Building something in health?
If you're working on care delivery, clinical tools or medical education: let's talk.