About KAVRIQ

KAVRIQ is a technical publication for engineers who want to understand the science behind AI.

Much of today’s AI content focuses on tools, demos, and surface-level abstractions. KAVRIQ takes a different approach: explaining the mathematics, models, and systems behind modern AI from first principles.

The goal is not simply to keep up with the latest framework, but to understand how things work underneath: the probability, optimization, representation, reasoning, and engineering ideas that shape AI systems.

About Me

I am Ravi Shankar, a software engineer working across AI systems, developer platforms, and cloud infrastructure.

I graduated in Computer Science from MNNIT Allahabad and have worked across startups, product engineering, cloud infrastructure, and large-scale platform environments. My experience includes THB, Urban Company, Flynote, Trinkerr, Oracle, and Salesforce.

KAVRIQ is where I document what I am learning, building, and trying to understand. My current areas of study include machine learning, deep learning, agentic AI, retrieval, evaluation, AI infrastructure, and the mathematical ideas behind these systems.

I care deeply about clarity, depth, mathematical intuition, practical judgment, and the distance between an impressive demonstration and dependable engineering.

Why KAVRIQ Exists

I am the son of a teacher, and teaching has always held a quiet place in my imagination.

Growing up, teaching was just one of many ambitions. I also wanted to become a physicist, mathematician, author, entrepreneur, and someone who built ambitious things. Over the years, I pursued projects, built products, opened a small coaching center in Greater Noida while working full-time in Gurugram, and eventually left my job to prepare for the UPSC civil-services examination.

Some of those attempts failed. Others simply did not become what I had imagined. Beneath many of them was a desire not only to do meaningful work, but also to be known for it.

In trying so hard to become something, I gradually lost touch with people I had met along the way: college friends, schoolmates, and former colleagues. Human connection, something I cared about deeply, became one of the casualties of my ambition.

I look for different things now: simplicity, human connection, a sense of community, and work that becomes more meaningful with time. I have become more aware that a life is finite. Not every ambition can become a life’s work, and not every effort needs to become an identity. I would rather be known by a small circle of people for something honest, and remain connected to them, than chase a larger idea and become a machine.

KAVRIQ emerged from that realization.

It gives me a place to learn in public, teach what I understand, and remain connected to people who care about understanding and building things well. It also gives me a constructive way to deal with loneliness—something I have increasingly felt while living and working in Bangalore, and something I suspect many other professionals experience too. Articles allow me to work through difficult ideas. Conversations and podcasts allow me to learn from others while making their knowledge useful to a wider audience.

Teaching and learning give KAVRIQ its form. The desire to understand things deeply gives it direction. Human connection gives it its purpose.

KAVRIQ may be a technical publication, but it is built with a human need for connection.

KAVRIQ may never become a large identity. A small, honest one would be enough.

A Working Philosophy

Understanding how things work is more valuable than chasing tools.

The goal is not to keep up with trends, but to build a mental model that lasts: mathematical where needed, practical where useful, and honest about uncertainty.