Available for interesting problems
Fullstack developer, app tinkerer & security enthusiast.
➜ ~ whoami
Ajay P Sakthikumar
➜ ~ cat stack.txt
fullstack · security · ai/llm
➜ ~ ls history/
flyfin-ai/ harness/ sap-labs/ nit-c/
➜ ~ ▍
./about
I care about clean code, delightful UX, and building things I'd actually use myself.
I build across the stack — from Next.js frontends and Python/Django backends to Electron desktop apps and offensive security tooling.
Most recently I worked on AutoTax at FlyFin AI, an AI-powered US tax preparation product. Before that, I was a Security Research Engineer at Traceable (acquired by Harness), running pentests for enterprise clients like Walmart, Morgan Stanley, and Disney, and shipping tooling that halved analysis time.
I started my career at SAP Labs India, where I automated CRM regression, built GPT-driven prototypes for service workflows, and co-authored a filed patent.
End-to-end features across UI, APIs, data, and desktop — with an eye on DX.
GPT-powered products, RAG pipelines, and intelligent workflow automation.
Pentesting, offensive tooling, and shipping fixes back into the product lifecycle.
./timeline
Milestones from campus to today — each stop shaped how I build and learn.
FlyFin AI
Bengaluru, India
Worked end-to-end on AutoTax — FlyFin's AI tax preparation product — across the Next.js frontend, Python/Django backend, and Electron desktop app.
Traceable Software India (acquired by Harness)
Bengaluru, India
Built custom security testing tools and led penetration tests across services for enterprise customers.
SAP Labs India
Bengaluru, India
Worked on SAP S/4 CRM Service Management — automation, tooling, and GPT-driven service workflow prototypes.
SAP Labs India
Bengaluru, India
Summer internship focused on test automation and internal tooling for CRM modules.
National Institute of Technology, Calicut
Kozhikode, India
Graduated with an 8.16 CGPA. Coursework spanned systems, algorithms, ML, and security.
./projects
A mix of personal experiments, course projects, and one patent.
Personal Project
Jun 2025 — Present
A modular Agent-Based Modeling framework in Python with JSON-composable scenarios and reinforcement-learning-style reward loops.
May 2025
AI assistant for legal guidance (Indian Penal Code) built on a RAG pipeline with Ollama, Flask, and Qdrant.
Course Project · Apr 2022
A framework wrapping Metasploit for mass exploit execution across a network, with automatic shell handling on compromised systems.
Event · May 2020
Conducted an entry-level CTF hosted on CTFd with ~60 participants, featuring internally hosted challenges and shell simulations.
Course Project · Jan 2020
Built an experimental OS in C, implementing core kernel functionality from the ground up.
./skills
The stack I reach for across product work and side projects.
./certificates
A few of the specializations and courses I've completed along the way.
