Chalk in one hand, code in the other. I teach mathematics and build AI learning platforms: some begin as my own experiments, some serve the whole university. I'm interested in how AI can make teaching better without taking the teacher out of it.
Dr. Radhakrishnan V K · Assistant Professor (SG) of Mathematics
Sri Chandrasekharendra Saraswathi Viswa Mahavidyalaya (SCSVMV) · Kanchipuram
Mathematics has been my classroom since 2009: a year at Sri Sankara Arts and Science College, then SCSVMV, Kanchipuram, from 2010. I teach Operations Research, Graph Theory and Mathematical Modelling, the mathematics of how networks connect, how queues move and how decisions get made.
My doctoral work began close to home. Kanchipuram is a city of temples, so I studied it as a network: the routes that link its temples and the queues pilgrims stand in, using network optimisation and Monte-Carlo simulation.
Today the question I keep returning to is teaching itself. Students now arrive with AI in their pockets, and I want to know what it does to learning. So I try ideas in code: platforms built for my own students, some of which grow into tools for the whole university. Along the way I write down what works and what doesn't.
I design, code and deploy each of these myself, for real courses and real users. Some started as my own projects, out of curiosity; others were built for the university. They are where I test ideas about AI in teaching.
Starting with year-wise B.E. mathematics and growing to take in every mathematics course: topics that unlock in order, properly typeset mathematics, an exam engine and certificates. The AI tutor answers students only from their own course material and cites its sources. Started as my own project; now being extended across the university.
eduai-platform-tau.vercel.appPrompting, custom Gemini Gems, NotebookLM, AI-built interactive teaching tools, LaTeX workflows and research tools, with a session on hallucinations and ethics. No coding background needed.
ai4academics.web.appA phone-installable app that keeps session plans and teaching material across programmes, with an AI planning assistant and approval-based release of material to parents. Began as a way to manage my own classes at T.I.M.E., the coaching institute, and grew from there.
sessionbook.web.appA university-wide course with versions for each programme. Five units of five modules; each module opens only after the student scores 6 out of 10 on the one before.
paryavaranam.web.appSchedule, session-wise feedback, two-way attendance and verifiable participation certificates for the university's first-year induction. Built at the request of the first-year engineering coordinators.
kanchiuniv.ac.in/sipEight stages from number sense to advanced reasoning, with a checkpoint assessment for every submodule and QR-verifiable certificates. The first platform I built.
placement-aptitude-lms.web.appPlanned versus delivered lectures, timetables, leave with automatic detection of affected classes and delegation, and an analytics view for administration. Built on my own initiative; the plan is to bring it to the university together with EduAI.
The full learner lifecycle, from enquiry and admission to fees, ID cards, study material, contact classes and results, with a mobile app for students. Built for a friend's start-up project.
An AI tutor should teach what was taught, the way it was taught. Mine answers from the course material and shows its source.
Students move to the next module when they have understood this one, not when the calendar says so.
Marks, attendance and records are calculated exactly or entered by a person, never generated by an AI model.
Most students open everything on a phone, so every platform is built and tested there first.
Certificates can be checked online by anyone, so nobody has to write to an office to confirm them.
The tools I teach colleagues need no programming. The teacher stays in charge of the content.
Short notes on using AI in ordinary teaching: explaining a concept, setting a paper, checking understanding, and saving time on paperwork so there's more of it for students.
I work on Operations Research and Graph Theory, usually starting from a real system: a pilgrimage route, a railway junction, a hospital roster, a wholesale market.
Development of Optimal Network Model for Tourism and Queueing Management in Temples with Case Studies on Kanchipuram. Supervised by Prof. Dr. T. Venugopal, 2018.
I'm always glad to exchange ideas with teachers, researchers and people building for education.