Mission
Dogma is building the world's first belief-aligned reviewer AI—a "meta-model" that sits above today's large language models to audit, score, and filter their outputs through the lens of values, principles, and worldviews.
Just as Geoffrey Hinton has suggested AI must one day learn instincts like care and responsibility, we believe the future of AI requires alignment not only with goals, but with beliefs. Human knowledge systems—philosophy, religion, culture, and mathematics—have always provided societies with formal frameworks for truth, ethics, and meaning. Dogma brings this lineage into the AI era by encoding these frameworks as transparent, configurable layers of abstraction that guide how AI interacts with humans.
Our Vision
Our vision is a world where every AI output can be understood, trusted, and critiqued through multiple belief systems—from ESG and DEI to Christianity, Islam, or Stoicism. In doing so, Dogma enables businesses, governments, and individuals to choose AI that reflects their values, fostering pluralism, accountability, and safety in an increasingly automated society.
At its core, Dogma seeks to mathematically formalize the audit of meaning—transforming subjective values into structured, reviewable signals. By acting as an independent referee across models, Dogma doesn't compete with LLMs—it ensures they remain faithful, interpretable, and aligned with the diversity of human beliefs.
Team
Nishith Krishna
Creator
Nishith Krishna is a Computer Science major from Indian Institute of Technology (IIT) and a Henry McCracken Fellow from the Courant Institute of Mathematical Sciences, New York University (NYU) ranked among the top three in the world for Applied Mathematics. He has authored multiple research papers on meaning of consistency large-scale distributed systems during his work at Microsoft Research Cambridge.
Nish earned his MBA from the Darden School of Business and thereafter led lending and payments businesses at Capital One, M&T Bank, E*Trade, and PenFed, where he scaled a new consumer lending unit to $12M ARR in a year after inception. He later founded Konduit, a Techstars and VC-backed consumer data exchange that processed over 10 million API calls per month.
Prajwal Nayak
Co-creator
Prajwal Nayak is a Computer Science graduate from Manipal University and a former Senior Software Engineer at Swiggy, a $11 billion Indian tech startup where he built a multi-stage canary deployment system for AWS Lambda and Step Functions, adopted by 200+ services across 40 teams.
In 2024, he won 2nd place at MetaAI's Llama Impact Hackathon in San Francisco and was the sole recipient of a full scholarship among 2,000 startup founders for the elite Hero Program at Draper University, founded by veteran venture capitalist Tim Draper. His work includes published research on ICU patient length-of-stay prediction and a widely-read five-part AWS Serverless Computing blog series.