Cogito, Ergo Sum
I build systems that learn—neurons in vitro, algorithms in silico, and habits in life. The aim is simple: move biology and people toward healthier dynamics.
About
I’m Nikhil—engineer, researcher, and builder. I’m exploring closed-loop neurostimulation on microelectrode arrays, computational models like an Epileptor with chloride dynamics, and a “Virtual Cell” for gene-expression prediction. Parallel to science, I design life systems. The thread through it all is feedback—observe, adapt, and improve.
“Radically self observe. The most important changes in your life will come from brief moments of deep introspection. Make your updates generous, and your corrections even more so.”
Selected Projects
SafeMODRL–Epileptor
Multi-objective RL over an Epileptor with chloride dynamics; phase-locked stimulation and GP safety wrappers for seizure suppression.
- Gym-compatible environment & Pareto archive
- Signal features + policy evaluation
Director and Found of NIP - Closed-loop MEA Platforms
Low-cost, reusable 16-channel bidirectional MEA system (STM32/PIC) for synaptic plasticity experiments and rapid CNS screening. Website:
Virtual Cell (In-Silico)
Predicting transcriptomic shifts under gene knockouts using DepMap + Perturb-seq; evaluation against benchmarks and ablations. Git:
Financial Systems
Trading habits that reduce friction, increase consistency, and improve risk management.
Notes & Writing
- Systems Thinking
How feedback loops craft better behavior.
- NeuroSymbolic AI - Our Future
Balancing invention with responsibility in neuro-AI.
Contact
Want to collaborate on neuro-AI, MEAs, or systems design? Reach out:
Email: iyer0106 [at] umn.edu