herentug

Built an on-chain protocol for efficient staking reallocation on Solana using Rust. Implemented PDA-based bookkeeping and staking account lifecycle management. Developed an algorithm to minimize on-chain transactions by matching off-setting user requests.

Led a 7-person AI team and scaled ML-driven staking and DeFi analytics products, contributing to a company valuation of $75M. Designed predictive models for validator performance, risk-adjusted yields, and cross-chain strategy optimization. Directed research and engineering roadmap in AI-driven protocol design, quantitative risk modeling, and automated on-chain decision systems.

Built an on-chain implementation of StableSwap-based debt calculation for DeFi lending and collateral systems. Implemented ML-based LST risk modeling for tokens such as stETH, rETH, cbETH, eETH, and swETH, estimating price divergence, depeg risk, liquidity risk, and yield efficiency using forecasting, probabilistic models, and Bayesian inference. Designed scalable on-chain infrastructure for treasury health monitoring, liquidation risk, and collateral optimization.

Developed early oracle infrastructure, including on-chain oracle contracts, multi-sig governance flows, and staking monitors. Built real-time monitoring tools for validator performance, slashing risk, reward estimation, and protocol telemetry. Contributed to early smart contract development and protocol MVPs across staking, governance, and treasury infrastructure.

Developed high-throughput asynchronous pipelines for Ethereum staking operations using web3.js and web3.py. Built and maintained on-chain oracle systems to aggregate and broadcast validator performance and staking metrics. Optimized block-level data retrieval and processing, reducing latency and improving overall staking efficiency.

Developed autonomous economic agent networks for supply-chain information exchange using ERC-20 token FET. Implemented and tested ML algorithms within smart contracts for synchronous data handling.