THE FOUNDATION & R&D
➔ A world where academic research is accelerated by autonomous AI swarms.
➔ Educational systems that adapt dynamically to personalize learning for every student.
➔ Secure and trustworthy intelligence networks resilient to post-quantum threat vectors.
We stand at the precipice of cognitive automation. Traditional development constructs are being rewritten by autonomous multi-agent intelligence.
Currently pursuing Computer Science at Symbiosis Institute of Technology, Pune, my vision is to construct secure, self-healing, post-quantum agent networks designed to augment complex enterprise decision swarms.
"We do not study history. We build systems that rewrite it." Leveraging advanced Artificial Intelligence, financial modeling systems, and deep cryptographic foundations to construct future companies, products, and secure networks that impact global scaling limits.
AI Bias Evaluation & Prompt Safety
Engineering evaluation benchmarks to analyze context-boundary leakage in Large Language Models. Proposing real-time mitigation prompt injection matrices.
Manuscript in prep // Bias in LLMsPost-Quantum Cryptography for Multi-Agents
Formulating quantum-resistant algorithms to secure inter-agent communication channels against interception in next-generation computing swarms.
Manuscript in prep // Lattice-Based Agent NodesGesture-Controlled Human-Computer Interfaces
Architected IntentCalc, a touchless virtual calculator driven by OpenCV and MediaPipe. Employs edge-triggered pinch controls, swipe backspaces, and fist-clears.
OpenCV // MediaPipe // Gesture AI// THE QUESTIONS I'M CHASING
Can AI discover novel research directions?
Can education become truly personalized?
Can small teams outperform large organizations with AI?
How do we build trustworthy intelligence systems?
THE MANIFESTO
OBSESSIONS VS ANTI-GOALS
// OBSESSIONS
// ANTI-GOALS
CONTRARIAN BELIEFS
➔ Build before you specialize
Specialization in a classroom is dead theory. True expertise is born from building systems that break under real-world conditions.
➔ AI will amplify curiosity, not kill it
Those who fear automation are doing work that deserves to be automated. High-agency minds will leverage AI to pursue 10x deeper questions.
➔ Small AI teams beat massive enterprises
Empowered by cooperative multi-agent swarms, a team of 3 builders can now construct, scale, and distribute systems faster than a 100-person enterprise committee.
THE BET
"I believe the next generation of intelligence systems will be built by small teams empowered by AI, not giant organizations."
Everything on this website is a step toward proving that belief.
ACTIVE VENTURES
Not projects. Emerging architectures built to solve cognitive friction.
ScholarMind
Automates academic literature reviews, identifying research gaps and drafting manuscripts. Orchestrates 6 specialized research agents using a self-healing LangGraph state machine and React Flow graph database representation.
MarketMind
Local-first intelligence system simulating 10 analytical investment agents. Employs dynamic DCF/WACC valuation math models, interactive accountability ledgers, and real-time portfolio analytics.
Symbi Club Control
ONGOINGLMS-inspired enterprise club coordinator for Symbiosis Institute. Manages permissions, event approvals, custom member registries, and secure cryptographic hierarchies across campus divisions.
Virtual Museum Guide
Selected to the Grand Finale of the Google Student Ambassador Hackathon 2026. Delivers highly-contextual historical RAG companion dialogues using multi-agent coordinate indexing.
PhishGuard
Machine learning-driven URL filter securing networks from social engineering vectors. Secured 3rd place with Team TensorSmiths at the SIT Ideation Summit.
VisionAct
Advanced UI automation engine that interacts with applications using direct visual context, enabling complex workflow automation where APIs or standard element trees are completely unavailable.
ACTIVE EXPERIMENTS
[01] RUNNING
[02] EXPLORING
[03] CONSIDERING
PUBLIC PREDICTIONS
AI research workflows will become partially autonomous, shifting from co-piloting to self-directed swarms.
Universities and campus divisions will increasingly adopt customized, peer-modeled personalized learning networks.
Multi-agent autonomous systems will become industry standard knowledge workers, replacing monolithic structures.
TELEMETRY & SECURE DATA
Real-time compute telemetry coupled with secure R&D credentials.
Selected into a highly competitive industry internship cohort focused on Artificial Intelligence, Advanced Software Development, and expert Mentorship. Earned flagship Internship Certificate.
Developed production-level LLM and Prompt Engineering pipelines, working with transformers, AI workflows, Python and Machine Learning architectures. Received outstanding remarks, complete letters of recommendation, and internship validation.
Contributing to the core R&D of **Symbi-GPT** under Dr. Ketan Kotecha. Designed and deployed advanced Retrieval-Augmented Generation (RAG) pipelines integrating vector stores, LangChain, and PostgreSQL.
Completed advanced risk assessment modules simulating institutional asset portfolios. Developed analytics evaluating hedging parameters and market risks.
Participated in mock enterprise architectures, debugging, refactoring, and code review processes aligned with Scandinavian agile teams.
Simulated consultant-level data analysis on Forage. Formulated analytics and dashboard reporting metrics evaluating corporate databases.