When to use machine learning
A practical way to choose between rules, pretrained models, and custom machine learning.
Dheeraj Inampudi
I'm an engineer who became a founder, which mostly meant learning that the hard problems stop being technical. I write about that here: the decisions, the trade-offs, and the frameworks I wish someone had handed me earlier.
Currently building EigenH AI.
Recent writing
Read the BlogLeadership principles I practice
Notes on clarity, ownership, and leading engineering teams when code is cheap and judgment is not.
LLM architecture mental models
How context, attention, activation cost, and failure behavior shape model selection.
What I write about
AI systems
Models, agents, context, evaluation, and production decisions.
Entrepreneurship
The decisions, habits, and tradeoffs involved in building a company.
Startup growth
Finding focus, growing teams, and keeping execution close to the customer.
Helping founders
Lessons and practical questions for people building their first company.
Leadership
Teams, ownership, technical judgment, and the work around the work.
Founder notes
Decisions, mistakes, and lessons from building a company.
Healthcare AI
What responsible automation looks like inside real practices.
Spaces
Browse allNow
I’m building EigenH AI with a focus on patient-access work for independent healthcare practices. I’m also thinking about how founders grow companies, how teams make decisions, and where AI systems create useful leverage without taking judgment away from people.
About me
My background moves through applied mathematics, data science, product engineering, and engineering leadership. The common thread is an interest in how technical ideas become useful systems and how teams learn to build them well.