Rohan Kulkarni
I build AI systems for health, from research pipelines to products used by 30K+ people.
- Co-founder and CTO of Fitfo: mobile, backend and LLM coaching for 30K+ people
- Speech and LLM pipelines that turn clinical sessions into insurance-ready notes
- ML on biomedical data: 50K+ ECGs, single-cell lung fibrosis, CRISPR base editing
B.S. Computer Science & Bioengineering at Illinois. Research across four labs in CRISPR oncology, nanomedicine, cardiac ML and computational biology.
Selected work
More projects
An AI decision engine for TikTok Shop affiliate creators: one action brief a day, then double-down or kill verdicts against their own numbers.
React Native + FastAPI app that estimates Rate of Perceived Exertion from workout video with pose-based CV, validated on 200 videos and 1,000 reps.
PyTorch CNN trained on 10,000+ labeled leaf images, about 87% top-1 with augmentation and class-imbalance handling.
A menu-bar task board for macOS with Linear-style statuses in a floating panel, synced to Supabase.
Experience
- Apr 2026 – Present
- Built and shipped LLM-based agentic systems that let creators deliver personalized coaching to thousands of users.
- Scaled Fitfo to 30K+ users and $200K+ ARR, owning system architecture and leading a team of 8 engineers.
- Lead the React Native, FastAPI, PostgreSQL and Stripe stack, plus the cloud infrastructure behind creator launches.
- May 2026 – Aug 2026
- Characterized epithelial cell states in single-cell RNA-seq data from a bleomycin-induced lung fibrosis mouse model.
- Implemented an end-to-end Seurat pipeline for QC, normalization, clustering, UMAP and marker-gene annotation.
- Ran differential expression and pathway analyses to identify regulators of lung epithelial repair and injury response.
- Sudden Cardiac-Death Awareness Research FoundationAug 2025 – May 2026Machine Learning
- Built preprocessing and data-quality pipelines for 50K+ ECG recordings, reducing labeling inconsistencies by 40%.
- Trained TensorFlow arrhythmia classifiers that improved cardiac-arrest detection by 12% AUC over the baseline.
- Analyzed subgroup errors across 10K+ patients to improve the models' demographic generalization and reliability.
- UIUC Department of Nanotechnology & NanomedicineSep 2025 – May 2026BionanomoleculesNie Lab
- Improved nanoparticle–biomolecule binding efficiency from 60% to 85% across 30 conjugation trials and lab assays.
- Quantified nanoparticle uptake in 1,000+ cells under 10 conditions, measuring 1.8× higher internalization in vitro.
- Reduced assay failure rates from 30% to 10% through 4 successive rounds of protocol validation and optimization.
- Sep 2022 – Jun 2024
- Designed inducible CRISPR base-editing workflows in mouse organoids, raising on-target editing efficiency to 70%.
- Validated edits in multiple organoid models, improving reproducibility while reducing assay reruns to below 10%.
- Analyzed tumor neoantigen and immune-response data from 50+ samples to prioritize new oncogene perturbations.
Education
Coursework
Activities
- Illinois Solar Car · ECE Engineer
- Illinois Business Consulting
Skills
About
Hi, I'm Rohan Kulkarni — a computer science and bioengineering student at Illinois, and co-founder and CTO of Fitfo.
I'm interested in using computation to make healthcare more predictive, personalized and data-driven, and I'm drawn to systems that turn complex biological data into real clinical impact. Alongside Fitfo I've worked across four labs: CRISPR base editing at Weill Cornell, nanoparticle uptake at Illinois, arrhythmia models on 50K+ ECGs, and single-cell analysis of lung fibrosis at Mount Sinai.