Graft processing unit
Receives extracted grafts, then grades, sorts and counts them automatically in a chilled, hydrated tray.
COSMIUM · FOLLI-PRO
Every graft, accounted for.
Folli-Pro brings computer vision and automation to the most manual stage of a hair transplant: processing and implanting thousands of grafts by hand.
In development · Not yet approved for clinical use
Built with clinical mentors from AIIMS New Delhi · Supported by KIIT-TBI and IIT Indore
SUPPORTED BY AND WORKING WITH
- Department of Biotechnology,
Govt. of India
THE PROBLEM
A precision procedure, still run on pen and paper.
In a typical hair transplant, around 3,000 grafts are extracted, sorted, counted and implanted, almost entirely by hand. Results depend on how skilled and how tired the team is, and the record of what happened lives on a scrap of paper.
-
8–12 h
procedure time
for an average of 3,000 grafts
-
6–8
trained technicians
in addition to the surgeon
-
~6 h
graft survival outside the body
every hour of handling counts
The result: unpredictable outcomes, long surgeries and no objective measure of quality.
THE SOLUTION
Meet Folli-Pro.
Folli-Pro automates the critical steps of graft processing and implantation, turning a skill-dependent manual workflow into a standardized, data-driven one.
In development · Not yet approved for clinical use
-
Computer-vision quality analysis
Identifies high-quality grafts from their anatomy.
-
Automated sorting and counting
Robotic segregation of grafts, with no manual tallying.
-
Optimal storage environment
Designed to hold grafts at 2–4 °C and keep them hydrated.
-
Real-time statistics and reports
Count and type for each and every graft, in a report for the surgeon and the patient.
-
Automated implanter loading
Back-to-back implantation from a pre-loaded cartridge.
Features describe the system under development. Performance has not yet been clinically validated.
Multi-graft implanter
Loads automatically from the processing unit and places grafts back to back, without reloading between grafts.
HOW IT WORKS
Three stages. One standardized workflow.
-
01 Extraction
The surgeon extracts follicular units exactly as today. Folli-Pro fits into the existing procedure.
-
02 Processing
Grafts go into the processing unit, where they are graded by computer vision, sorted, counted and held chilled and hydrated.
-
03 Implantation
The multi-graft implanter is loaded automatically and places grafts back to back.
Example from a single procedure: 1,800 + 800 + 600 grafts = 3,200, counted automatically by type.
THE TECHNOLOGY
Seeing what the human eye misses.
Each graft is only about 0.8 × 5 mm. Folli-Pro's computer-vision model looks for the anatomical structures that indicate how likely a graft is to survive, and grades every graft before it is implanted.
Grade 1High survival22
Grade 2Low survival9
Grade 3No survival5
Illustration of automated grading and sorting.
14,000+
graft images collected for model training, under ethics approval at AIIMS New Delhi
1,200 / hr Design target
target implantation rate with back-to-back loading
WHY IT MATTERS
Better for surgeons, clinics and patients.
Our design targets, to be validated in clinical investigation:
Standardized quality
Objective grading for every graft, for more predictable outcomes and greater patient trust.
About 6-hour procedures
Targeting around 30% less surgery time: less patient discomfort, better graft survival.
Teams of 3
Targeting around half the technicians, with less variability between operators.
Full transparency
Real-time insights and a procedure report for the doctor and the patient.
How Folli-Pro compares
| Manual tools | Semi-automated systems | Robotic systems | Folli-Pro | |
|---|---|---|---|---|
| Graft sorting and counting | Manual | Manual | Manual | Automated |
| Real-time graft analytics | None | Total count only | Limited | Count, quality, transection |
| Graft quality analysis | None | None | None | AI-based assessment |
| Implanter loading | Manual | Manual or suction-based | Manual | Automated, suction-based |
PROGRESS
From the operating room to a working prototype.
Observed in the operating room
Observation in working hair transplant clinics to map every manual step.
Validated with 50+ clinicians
Plastic surgeons and dermatologists across Delhi NCR and at national conferences.
Provisional patent filed
Indian patent application No. 202511111643.
Ethics approval granted
For data collection and AI model training at AIIMS New Delhi.
14,000+ graft images collected
Captured during live procedures.
Computer-vision model trained
Detects graft anatomy and assigns quality grades.
Proof of concept V1
A working graft-processing prototype.
User studies and doctor feedback
Hands-on sessions with clinicians that shape each iteration.
Now in progress
- Final prototype: graft processing plus implanter
- Complete patent filing
- Conclude the ethics-approved research study
ROADMAP
A clear path to the clinic.
- Device class: Class B (India)
- Regulator: CDSCO
- Quality system: ISO 13485
- International: CE Mark, then US FDA
Planned milestones. Timelines may change as development and regulatory review progress.
SUPPORTED BY
Backed by India's medtech ecosystem.
MeitY Genesis EIR
through KIIT Technology Business Incubator, Bhubaneswar
₹10 lakh
Startup India Seed Fund Scheme (SISFS)
through IIT Indore
₹20 lakh
Clinical mentorship
Department of Burns & Plastic Surgery, AIIMS New Delhi
Technical mentorship
Centre for Biomedical Engineering, IIT Delhi
Innovation fellowship
School of International Biodesign, AIIMS & IIT Delhi
Department of BiotechnologyGovernment of India
TEAM
Engineers and clinicians, building together.
Anurag Singh
Founder
5 years in industry · Ex-Oracle · Electronics engineer · Innovation Fellow at AIIMS
Rishab Gupta
Embedded Engineer
5 years · B.Tech, Electronics & Communication
Sarang Banakhede
AI/ML Engineer
2 years · B.Tech, Computer Engineering
Vikas Kumar
Product Designer
1 year · B.Des, Industrial Design
Mentors
Dr. Maneesh Singhal
Clinical mentor
Professor and Head, Burns & Plastic Surgery, AIIMS New Delhi · MS, M.Ch, FACS, FRCS, MNAMS
Dr. Shivangi Saha
Clinical mentor
Assistant Professor, Burns & Plastic Surgery, AIIMS New Delhi · MS, M.Ch, MNAMS
Dr. Deepak Joshi
Technical mentor
Professor, Centre for Biomedical Engineering, IIT Delhi · PhD
Nitin Sisodia
Business mentor
Founder, Sohum Innovation Labs · Product designer, NID Ahmedabad
WORK WITH US
Pilot Folli-Pro in your clinic.
We are partnering with a small group of high-volume hair transplant clinics in Delhi NCR to shape Folli-Pro before launch. We are also glad to hear from investors, researchers and engineers.
- hello@cosmium.co
- Based in
- New Delhi, India