We do research-inspired software development.
Born from a research organization in mathematical sciences, Panini Eight builds software the way researchers solve problems—with deep domain understanding, rigorous thinking, and creative insight.
What does research-driven software mean?
Software where new knowledge creation is part of the product—not just assembling tools, but understanding problems deeply enough to discover something new.
AI and ML
Training new models and reasoning systems, not just calling APIs
Mathematical Computing
Pure mathematics, statistics, and formal verification applied to real systems
Domain-Deep Engineering
Studying the problem with the rigour a researcher brings to a field of inquiry
Adaptive Systems
Software that learns, optimises, and evolves in education, healthcare, and operations
Software shaped by deep understanding
At the intersection of research methodology and production engineering.
Web Applications
Full-stack applications built around rigorous domain models. Complex workflows, role-based access, multi-tenant architectures, and systems where correctness is non-negotiable.
AI and Machine Learning
Research-grade ML pipelines and intelligent systems. From computer vision to mathematical reasoning engines, AI grounded in theory and validated through experiment.
Domain-Driven Consultation
Technology audit and operations analysis with a researcher’s eye. Architecture review, security audit, and creative solutions at the structural level.
Projects that demonstrate our approach
Lokeswarananda Eye Foundation—Research-Focussed Hospital Management System
A next-generation HIMS that transforms hospitals into engines of scientific discovery. Combining interoperable standards, clinical compliance frameworks, and India’s digital health infrastructure into a unified platform.
Panini8 Olympiad Software
AI-driven adaptive practice for Cheenta Academy students. Timed mock tests, real-time progress reporting, doubt clearing, and intelligent problem selection calibrated to competition standards.
Filibot—Competency-Based Learning Automation
Automates the implementation of competency-based learning mandated by India’s National Education Policy. Helps schools transition from rote assessment to structured competency mapping, tracking, and reporting—making NEP compliance practical and measurable at scale.
Where mathematical research meets artificial intelligence
Three focus areas: AI-assisted mathematical research, mathematics-assisted AI research, and AI resource development for teaching mathematical sciences.
AI-Assisted Mathematical Research
Exploring how Lean 4, Isabelle, and evolutionary algorithms can assist frontier mathematics. Our work includes studying the intrinsic geometry of formalised mathematical knowledge—the Lean Mathlib dependency graph (619,518 nodes) and its hyperbolic tree-of-cliques structure.
Read our papers →Geometry as Infrastructure for LLMs
Can geometric structure reduce hallucination in large language models? We research how Poincaré ball embeddings and curvature-aware architectures provide navigational structure for AI systems working with formal proofs.
Read the vision note →AI for Mathematics Education
Randomised control trials exploring how AI improves human collaboration in virtual classrooms. Comparing traditional assessment, AI-chatbot discussion, and AI-paired peer discussion for impact on curiosity and problem-solving.
Learn about the Bose study →Our research team
Researchers across geometric group theory, computer vision, quantum information, algebraic K-theory, and medical AI—from ISI, IISc, IISER, IIT, Penn State, and Auburn University.
Born from Cheenta Academy
Panini Eight is the technology arm of Cheenta Academy, a research and education organization in mathematical sciences founded in 2010. Depth over breadth. Understanding over imitation. Research as the most powerful form of problem-solving.
Cheenta Founded
Olympiad training with research-first pedagogy from day one.
Software Practice
Internal adaptive learning tools evolve into a full engineering practice.
Panini Eight
Research-inspired software for organizations worldwide.
Researchers who build
Backgrounds in mathematics, statistics, AI research, and engineering from ISI, IIT, and Penn State.
Dr. Ashani Dasgupta
Srijit Mukherjee
Raghunath J V
Have a problem worth solving?
We work best with organizations who value understanding over shortcuts. If your project needs deep domain thinking and rigorous engineering, we should talk.