Senior Infrastructure & AI Systems Engineer
High-Throughput Distributed Systems · Performance Engineering · Autonomous Agent Infrastructure

LinkedIn · Notion Portfolio · jintukumardas@gmail.com

I bridge the gap between low-level silicon efficiency and high-level autonomous intelligence. My career spans optimizing core math kernels on bare metal at AMD to architecting and scaling resilient backend infrastructures—spanning both modular distributed systems and high-throughput monolithic architectures—that process 500M+ transactions and manage networks securing 9B+ in assets.

Whether it is conquering complex concurrency race conditions, custom-building zero-copy serialization layers, or designing cryptographic spend-safes for autonomous AI agents, I build for predictable, deterministic performance under extreme production stress.

• 📈 Massive-Scale Data Planes: Designed, optimized, and managed web-scale backend infrastructure handling 500M+ transactions during peak protocol volume and anchoring critical transaction networks securing over $9B in capital.
• ⚡ Low-Latency Serialization Engineering: Architected custom binary serialization protocols achieving a 30% lift in data-plane throughput and slashing P99 latency on execution layers to ensure stability under heavy concurrent load.
• 🔄 Monolithic & Distributed Architecture: Proven capability re-engineering monolithic frameworks into modular, highly scalable distributed event-driven systems, successfully compressing integration cycles from weeks to days.
• 🔬 Academic Roots in AI/ML: Hold an Master's degree in Computer Science with an AI/ML focus, backed by 2 published research papers in data dimensionality and graph-based analysis with IEEE and Springer.
• 🛠️ Founder-Speed Execution: Prolific hackathon builder with multiple global wins across AI Agent systems, high-performance infrastructure, and automated transaction rails.

• Languages: Go · Rust · Solidity · C/C++ · Python · TypeScript · JavaScript · Java
• AI Infrastructure: Agentic Workflows · Model Abstractions · RAG · Context Engineering · MCP
• Systems & Infrastructure: Distributed Systems · High-Performance Networking · Concurrency Models · Order Routing · High Availability
• DevOps & Diagnostics: AWS · Kubernetes · Docker · CI/CD · perf · GDB · Valgrind

• Performance Engineering: System-level profiling, micro-architectural optimization (cache, memory alignment), SIMD vectorization, and fixing complex race conditions under high load.
• Backend Architecture: Shifting large codebases between monolithic architectures and robust distributed networks with sub-second delivery, automated fallbacks, and real-time state consistency.
• Agentic Workflows: Designing custom internal execution environments, fallback retry mechanisms, and production-grade security layers enforcing strict programmatic spending constraints for autonomous agents.

• 🔹 ByteMentor AI – Full-stack adaptive technical training platform featuring an interactive AI interviewer, 19 automated system practice modes (System Design, Security Audits), and real-time behavioral gap analysis. Built a modular LLM abstraction layer swapping dynamically between Claude, GPT, and Gemini.
• 🔹 LedgerMind – A production-grade escrow and payment intent architecture tailored specifically for autonomous AI agents. Enforces strict, human-signed spending constraints coupled with immutable, on-chain verifiable audit trails.
• 🔹 EON – An LLM-backed natural language interface allowing users to seamlessly interact with complex infrastructure protocols (token routing, asset minting, smart contract deployment) purely via contextual text conversations.
• 🔹 Giggle – A rapid chat-integrated payment gateway mapping consumer chat environments (WhatsApp) directly to financial settlement layers for instant, zero-friction settlement.

• 🥈 Vyper/Polygon AI Agent Hackathon (2025) – 2nd Place for building specialized autonomous agent logic 🔗
• 🏅 Circle Developer Bounties (2025) – Track winner for automated multichain payment rails 🔗
• 🥈 ETHGlobal Singapore – NounsDAO Expand Public Goods Winner 🔗
• 🌟 Committer of the Month – Shardeum (Recognized for core protocol performance optimizations).
• 🌟 Director Spotlight & Extra Mile Awards – AMD (Recognized for exceptional multithreaded algorithm and CPU math library optimizations).

• 📄 Reducing Dimension for Remote Sensing Image Dataset - IEEE, 2022
• 📄 A Graph-Based Band Selection Method for Hyperspectral Images Using Correlation Matrix - Springer, 2021

⚡ If you are engineering deterministic backend systems that need to process complex real-world state, orchestrate massive transaction load, or safely execute agentic intelligence at scale — let's build something relentless.