
2
2
5+ years of experience
I am Senior Software Engineer with over 6 years of experience in AI-first architecture, I focus on building autonomous systems that are both scalable and safe. My recent achievements include winning 3rd place at the 4th Global AI Hackathon (organized by MIT Club & Hack-Nation) and 1st place in the Agentic Commerce challenge via LabLab.ai and Google DeepMind. I don't just build models; I architect ecosystems. From leading teams at Icommunix to reduce onboarding time by 40% to prototyping zero-click payment SDKs for AI agents, I thrive at the intersection of technical depth and strategic automation. I am currently researching centralized intelligence for multi-model orchestration to create more resilient, decision-driven AI frameworks.

CipherScan AI is an enterprise-grade fraud investigation and trading intelligence platform that combines deterministic risk scoring with AI-powered analysis. The system detects fraud patterns through baseline analysis and deviation detection, while simultaneously analyzing trading behavior to warn traders about repeating bad patterns based on their own historical data. Key Features: Fraud Investigation: Real-time alert processing with AI-generated investigation narratives, severity classification, and complete audit trails Trading Intelligence: Autonomous behavioral analysis that detects emotional trading patterns and warns traders when they're repeating mistakes from their own trading history Enterprise Security: JWT authentication, AES-256 encryption, role-based access control, and compliance-ready audit logging Premium UI/UX: Noir fintech design with glassmorphism effects and intuitive workflows Technology Stack: Node.js, TypeScript, React, Google Gemini AI, SQLite Business Value: Reduces investigation time from hours to seconds, helps traders learn from their mistakes, and provides complete regulatory compliance with automated audit trails.
7 Feb 2026

OmniAgentPay gives AI agents a complete payment infrastructure layer. Think of it as Stripe for AI agents one SDK that lets any autonomous agent handle money safely and instantly. Traditional blockchain SDKs are built for humans with screens and confirmation buttons. Handing an AI agent a private key is a security disaster waiting to happen hallucinations, prompt injection, or simple logic bugs can drain a treasury in seconds. OmniAgentPay solves this. We wrap Circle's developer-controlled USDC wallets in an atomic Safety Kernel that enforces spending rules before any transaction executes. What agents get: Universal pay() Method: One call intelligently routes through direct USDC transfers, x402 pay-per-use APIs, or cross-chain CCTP transfers automatically. Atomic Spending Guards: Budget limits, rate limits, recipient whitelists, and transaction caps enforced atomically, even under high concurrency. Autonomous Cross-Chain: Full Agent-Side CCTP minting to Arc Testnet and other networks no external relayers required. Zero Blockchain Complexity: No ABIs, no gas math, no private keys. Just pay(). We're not building another agent application. We're building the infrastructure any AI agent can plug into to handle money whether it's LangChain, OmniCoreAgent, or a fully custom agent. Built With: Python, Circle Developer-Controlled Wallets, CCTP V2, x402 Protocol, Arc.
24 Jan 2026