Category: Generative — AI & ML

Project Description: AIlluminate is a suite of private AI tools packaged as offline-usable desktop applications with a focus on privacy, simplicity, and convenience. It’s an all-in-one desktop AI suite that does virtually anything the host machine can handle—everything from image generation and 3D modeling to code assistance and content creation—without ever sending data off-machine.

• Fully offline: all components run locally, ensuring zero data leakage.
• Cross-architecture support: custom binaries for Windows, macOS, and Linux.
• Modular suite: includes chat assistants, image/3D generation tools, narrative engines, and more.
• One-click installer: simplifies setup and automatically detects GPU/CPU capabilities.

My role: As lead architect and developer, I designed and built AIlluminate from the ground up. I wrote every component in a portable way so the suite would run on all machine architectures (x86, ARM, etc.), handled offline model packaging, and integrated each tool into a unified desktop interface.

Category: Infrastructure — AI & ML

Project Description: Bitcurrent is a distributed model‐training project built on the Bittensor network. It enables users to contribute compute power toward training AI models, and the generated earnings from training are automatically split among the node runners who participated in that round. The system emphasizes fairness and scalability, allowing for seamless collaboration and reward distribution across a decentralized mesh of contributors.

• Distributed training on Bittensor: nodes coordinate through a decentralized hash table (DHT) to exchange weights and gradients.
• Earnings redistribution: training rewards are algorithmically divided among all active node runners based on their contributed compute.
• Scalable architecture: supports dynamic node discovery and fault tolerance via a standalone DHT layer.

My role: I architected and implemented a more potent standalone DHT to improve node discovery and data exchange efficiency. Additionally, I integrated Petals-Lightning experimental protocols to further accelerate the training pipeline, reducing synchronization latency and boosting throughput across the network.

Category: Infrastructure — AI & ML

Project Description: Petals is an open-source distributed inference system for large language models (LLMs). By pooling donated GPUs into a decentralized network, Petals enables anyone to run inference for free, similar to a community-driven initiative like Folding@home. Users’ GPUs work together to host and serve large models, lowering the barrier to entry for AI experimentation.

• Community-first distributed inference: combines volunteer GPUs to host LLM checkpoints.
• Zero-cost access: end users can perform inference without paying for cloud resources.
• Scalable mesh network: dynamically discovers nodes and balances inference queries across the network.

My role: I created a series of inference demos featuring popular characters and packaged them as standalone desktop applications. These demos helped first-time AI enthusiasts get started quickly, showcasing Petals’ capabilities and simplifying the onboarding process for new users.

Category: Finance — AI & ML

Project Description: Circe AI is an automated trading platform that leverages AI-driven market analysis to execute trades at 30-minute intervals. The system maintains a 75% prediction accuracy rate and has consistently demonstrated at least 1% profit per day in live demos.

• 🔹 Automated 30-minute trading intervals powered by AI-based market forecasts.
• 🔹 Achieves a 75% accuracy rate in trade predictions under varied market conditions.
• 🔹 Consistent performance: reliably generates a minimum of 1% profit per day in demos.

My role: I designed and orchestrated the ensemble of machine learning models and runtime infrastructure that form the prediction engine. I handled the entire machine learning pipeline—from data ingestion and feature engineering to model development and integration—ensuring seamless prediction and trade execution.

Category: Security — AI & ML

Project Description: Testmachine.ai is a platform designed to find vulnerabilities in EVM smart contracts both prior to deployment and throughout the application’s lifecycle. It leverages bespoke AI pipelines to detect security flaws long before they can be exploited in production.

• Testmachine Scout processes thousands of EVM contracts per second using optimized AI-driven analysis for production and dataset-creation usage.
• Scout returns accurate vulnerability assessments along with a programmatically traversable intermediate representation.
• Scout performs cross-contract network analysis, checking interactions across multiple smart contracts.
• Scout generates “bad actor” contracts for automated exploit testing in a controlled environment.

My role: I designed and developed the Scout system. Scout can process thousands of contracts per second and output both a detailed vulnerability reading and an intermediate representation of each contract. It also supports network-wide analysis and automatically generates malicious contracts to test potential exploits in a safe test environment.

Category: Casino — Blockchain

Project Description: Funtoken (also known as Funhub) is the central entity governing the dPlay and freebitco.in casino platforms. As a Binance-listed asset, Funtoken operates a portfolio of bespoke casino games and services, primarily serving a Southeast Asia user base. The platform coordinates token economics, game deployment, and regulatory compliance across its ecosystem.

• Governance of dPlay and freebitco.in casino ecosystems under a single token brand.
• Binance-listed ERC-20 asset, facilitating broad liquidity and market access.
• Development and operation of multiple bespoke casino games tailored to Southeast Asian audiences.
• Integrated on-chain tokenomics to reward user engagement and drive platform growth.

My role: As CTO, I oversee all technological processes and strategic planning for Funtoken. I manage cost budgeting, ensure shareholder expectations are met, and lead a cross-functional tech team to deliver game development, blockchain integration, and platform maintenance.

Category: Smart Contract Game — Fantom Opera

Project Description: “Let’s Conquer Earth” is an on‐chain browser‐based strategy game on Fantom Opera. Players use the in‐game token to buy countries or pay to steal them from other players—each steal action generates yield for the current country holders. Every game action also drops collectible items; gathering enough items unlocks the Grand Champion role, granting that player a share of all house earnings. Deployed on Fantom Opera, the game generated $100,000 USD in volume within its first 24 hours.

• 🔹 Pay‐to‐steal mechanic: spend tokens to seize another player’s country, with fees distributed as yield to current holders
• 🔹 Item drops for every action: collect items from purchases, steals, and trades
• 🔹 Grand Champion system: accumulate items to become the sole Grand Champion, earning a portion of all house fees
• 🔹 Live on‐chain leaderboard and market demand-based country pricing that responds to game activity
• 🔹 Generated $100K in trading volume in the first 24 hours on Fantom Opera

My role: As CTO / lead engineer, I architected the entire dApp stack (smart contracts, front‐end custom game engone, CI/CD), performed security audits, and wrote gas optimizer scripts that reduced average gameplay gas by 28%.

Category: Digital TCG — Fantom Opera

Project Description: “The Lexicon” is a digital trading-card game on Fantom Opera. Players build decks of mythical cards, each with on-chain NFT attributes. The platform includes PvP matchups, tournament ladders, and a rarity algorithm that adapts supply over time.

• Fully on-chain pack and card minting using ERC-1155
• Dynamic rarity algorithm that shifts based on supply
• Integrated with MetaMask + WalletConnect
• Real-time trading tables and random p2p trade smart contracts

My role: Designed the smart-contract architecture, built the Bootstrap 5 frontend, and wrote the backend Moralis service that handles Oracle calls and off-chain deck statistics.

Category: Dev Tools — Multichain

Project Description: PAXL is a platform that automates the deployment and verification of EVM smart contracts on any supported network. By integrating with block explorer APIs and offering templated contract bundles, it slashes development time and cost for Web3 projects. Users can push a single command or click a button to compile, deploy, and verify contracts on Ethereum, BSC, Fantom, Polygon, and more—no manual verification steps required.

• Automated compile-and-deploy pipeline for any EVM-compatible chain
• Instant verification on block explorers (Etherscan, BscScan, PolygonScan) via integrated APIs
• Prebuilt contract templates (ERC20, ERC721, upgradeable modules) to bootstrap projects
• Multi-network support with configurable RPC endpoints and gas optimizations
• CI/CD integration hooks and CLI tool for seamless inclusion in developer workflows

My role: Architected the deployment automation engine (bun.js) that interacts with Solidity compilers and explorer APIs, designed endpoints for multi-network RPC orchestration, and built the bespoke-framework dashboard and CLI wrapper to let developers trigger deployment and verification with a single click or command.

Category: Blockchain Food Delivery — Polygon

Project Description: FoodChain Global is a decentralized platform that funds and coordinates food deliveries to regions struggling with access to nutritious meals. Donors purchase or stake the native ERC20 token to fund shipments, and every transfer is recorded on-chain for end-to-end transparency. Beneficiary organizations redeem tokens for supplies, ensuring aid reaches those who need it most.

• On-chain tracking of food shipments and distribution to ensure transparent delivery
• ERC20 token used to fund, stake, and redeem for food supplies in partner regions
• Pre-audited token contract passed third-party security audit seamlessly
• Responsive onramping dashboard built with Bootstrap 5 and sticky mobile layouts for seamless UX

My role: Developed and pre-audited the ERC20 token contract—enabling it to pass a third-party audit without issue—and built numerous responsive UI widgets using Bootstrap 5 with sticky mobile design to simplify donor interactions and on-the-ground coordination.

Category: Web3 Game — Polygon

Project Description: Nova Creed is an NFT-based web3 game on Polygon where players collect space-themed NFT avatars, upgrade them, and battle in PvP arenas. Each NFT has on-chain metadata for stats, and tournaments are governed by DAO voting.

• ERC-721 avatars with on-chain upgrade paths
• Polygon POS integration for low transaction fees
• On-chain DAO governance for tournament rule sets
• Real-time battle results via Chainlink VRF for randomness

My role: Architected the smart-contract system (ERC-721, ERC-1155, ERC20, and various custom uogradable logic contracts), built the React/TypeScript frontend, and designed the ship creation system with infinite expansion capability.

Category: Unity/Web3 — Multichain

Project Description: CrossChain Rally is a Unity-based multiplayer racing game integrated with Web3. Players can purchase on-chain vehicle NFTs, customize them, and race on procedurally generated tracks. The game supports cross-chain asset transfers between Chronos, BSC and Fantom Opera.

• Unity client with Web3 wallet connectors
• On-chain vehicle NFT minting and metadata storage
• Cross-chain bridge for transferring NFTs between chains
• Real-time race results recorded on-chain for leaderboards

My role: Developed the Unity-Web3 bridge using Web3 and Moralis, wrote the bridging smart contracts (LayerZero-based), and implemented the matchmaking and leaderboard backend in Node.js.

Category: Finance — Fantom Opera

Project Description: SAGA Stable is a next-generation stablecoin on Fantom Opera powered by an arbitrage vault that accepts multiple stablecoins as collateral. When a collateral token drifts too far below its peg, the vault automatically locks new inputs of that token—preventing further devaluation. This safety mechanism helps SAGA holders incur fewer losses compared to traditional stablecoins, as arbitrage opportunities are captured and collateral risk is minimized.

• Arbitrage vault aggregates USDC, DAI, USDT, and other collateral
• Dynamic input lock: disables deposits of any collateral falling >1% below peg
• Automated peg-stabilization through on-chain arbitrage triggers
• Reduced downside risk for SAGA holders via safety locks and diversified collateral
• On-chain dashboard displaying collateral health and vault status

My role: Architected and audited the arbitrage vault and collateral-locking logic, implemented the stablecoin mint/burn contracts, and built the Vue.js dashboard to monitor vault performance and peg stability.

Category: Finance — Fantom Opera, Polygon, Binance Smart Chain

Project Description: RNDM Token is a cross-chain ERC20 deployed on Fantom Opera, Polygon, and Binance Smart Chain that lets users stake directly with the token contract—no approve call required. Once per month, the contract randomly mints and distributes a 1% yield to active stakers, generating a randomized FOMO mechanic that keeps users buying and staking continuously.

• Cross-chain ERC20 implementation on Fantom, Polygon, and BSC
• Direct staking function built into the token contract (no prior approval needed)
• Monthly random 1% yield minting to a subset of stakers
• Randomized distribution drives ongoing FOMO and engagement
• On-chain staking dashboard showing active stakes and rewards

My role: Developed the core ERC20 staking logic, implemented the monthly randomized yield-mint mechanism, and built the React-based staking portal that interacts directly with the contract’s stake/unstake functions.

Category: Staking & Ecosystem — Fantom Opera

Project Description: This permissionless long-term staking system is designed to reward entire ecosystems by measuring dApp performance on-chain and algorithmically splitting earnings between the dApp owner and stakers of specific community assets. It enables any project to launch a staking pool without intermediary approvals, ensuring that rewards scale with actual usage and engagement.

• Permissionless staking: anyone can stake community tokens to support a dApp’s ecosystem
• Performance-based rewards: on-chain metrics feed into an algorithm that allocates revenue shares
• Earnings split: automatically divides protocol fees between the dApp owner and stakers
• Community asset pools: create separate staking pools per token
• On-chain analytics dashboard: real-time insights into staking volumes, reward rates, and dApp activity

My role: Architected and audited the staking smart contracts, developed the front-end interface in a bespoke JS framework, and built the performance analytics pipeline using Web3 libraries and smart contracts to track dApp usage and calculate reward distributions without massive overhead from indexing.