I build backend systems where correctness and security are first-class concerns, primarily within the Frappe ecosystem (ERPNext and HRMS). My work focuses on permission models, authorization, data lifecycle correctness, and designing software that fails safely rather than silently.
Alongside backend engineering, I develop LLM-powered applications, including Retrieval-Augmented Generation (RAG), LangChain, LangGraph, and agentic workflows, applying software engineering principles to AI systems.
- Backend Development — Python, Frappe, MariaDB
- Application Security — Authorization, Permission Models, Secure Design
- Applied AI — RAG, LangChain, LangGraph, AI Agents
- Open Source — ERPNext & HRMS Contributions
I approach software from both the maintainer's and the attacker's perspective. Every change is guided by the system's intended behavior, emphasizing correctness, reproducibility, and long-term maintainability over quick fixes.
Building systems that hold up under scrutiny.