Digital Research Archive
Abhishek Gupta
Formalizing the architecture of autonomous intelligence and adversarial defense. This ledger serves as a static record of verified engineering artifacts, research whitepapers, and high-stakes CTF methodologies.
Philosophy of Defense
Architecting Resilience
My work is predicated on the "Assumed Breach" model. I don't just build systems to function; I architect them to survive. By merging Adversarial Machine Learning with First-Principles Security, I develop autonomous agents capable of real-time self-audit and distributed recovery. From hardening infrastructure via MARL to protecting academic communities from recruitment fraud, my methodology prioritizes algorithmic consistency and formal verification over simple implementation.
Explore Live Architectures
My Journey
BCA — Computer Science
Enrolled at Babasaheb Bhimrao Ambedkar University. Foundation in Data Structures, Algorithms, and secure coding.
IBM Edunet — AI Intern
Developed AI-integrated web apps, optimizing page load times by 20%. First step into production AI.
Google AI Agent Intensive
Selected for intensive training on building AI Agents using Gemini & Vertex AI. 5-day program.
TCS HackQuest S10 — Round 2
Solved advanced Web Exploitation challenges and cleared intensive CTF rounds. Qualifier status.
Freelance AI Developer
Started on Upwork. Designing autonomous AI agents and API integrations for international clients.
Meta OpenEnv Hackathon
Architected multi-agent RL environment for Meta PyTorch Hackathon — autonomous security auditing.
Gemma-SOS — Edge AI Launch
Published fine-tuned Gemma 4 E2B model on HuggingFace. Built offline disaster response Android app.
Cyber Achievements
Meta OpenEnv Hackathon
Status: Round 1 Submitted
Multi-agent RL environment for infrastructure security auditing using PyTorch.
TCS HackQuest S10
Status: Round 2 Qualifier
Web Exploitation CTF — advanced challenges solved, intensive rounds cleared.
NPCI GFG #GeekStreak60
Status: 60-Day Streak
Unbroken 60-day streak solving complex DSA problems. Algorithmic consistency champion.
Google AI Agent Intensive
Status: Completed
Selected participant in 5-day training on building AI Agents with Gemini & Vertex AI.
Bug Bounty Hunter
Status: Active (P4/P5)
Actively hunting on HackerOne. Web vulnerability research and responsible disclosure.
Gemma-4 E2B SOS LoRA
Status: Published
Fine-tuned Gemma 4 model deployed on HuggingFace. Offline disaster response AI.