2
2
Kenya
3+ years of experience
AI-Oriented Software Engineer | Specializing in End-to-End Product Development I am a software engineer with a proven track record of architecting, building, and deploying scalable mobile solutions to production. My core expertise lies in developing robust systems from the ground up, with a deep focus on user-centric design and solving complex technical challenges. I am now channeling this end-to-end engineering rigor into the AI space, with a keen interest in building and orchestrating practical AI agents. I am adept at creating tools that enhance productivity and empower users, and I am eager to apply my problem-solving skills and production-minded approach to developing innovative solutions with IBM watsonx Orchestrate.

The Problem Crypto projects suffer from a "mercenary capital" problem. Investors often buy tokens, wait for a small pump, and sell immediately ("pump and dump"). There is currently no automated infrastructure on Qubic to identify and reward loyal investors who hold their positions for the long term. The Solution QXBoost is an automated "Diamond Hand" engine. It incentivizes long-term holding by turning passive investing into an active, gamified experience. It listens to the blockchain 24/7, identifies high-value investors, and rewards them automatically if they don't sell. How It Works Detection: A Make.com webhook monitors the QX Smart Contract. When a user buys >10,000 CFB, the system triggers. AI Engagement: Google Gemini AI analyzes the transaction and generates a unique, hype-filled alert ("🚨 Whale Detected!") sent to Discord and Telegram to excite the community. Retention Check: The system waits for a set period (e.g., 30 days) and then queries the Qubic API to verify the wallet still holds the assets. Reward: If verified, Gemini generates a custom HTML "Success Report," and the system emails the user a "Diamond Hand" NFT reward. Why It Matters By automating retention, QXBoost helps Qubic projects stabilize their token price and build stronger communities without manual effort.
7 Dec 2025

AutoExpenseFlow revolutionizes corporate finance operations by replacing manual expense processing with an intelligent AI agent built on IBM watsonx Orchestrate. The solution addresses the critical pain point of accountants spending hours manually reading invoices, validating expenses, and recording data across disparate systems. The agent autonomously processes expense documents through a sophisticated workflow: extracting vendor information, amounts, dates, and categories using advanced document understanding; validating each expense against company spending policies and compliance rules; seamlessly integrating with existing ERP systems like SAP or ServiceNow; and automatically generating comprehensive audit trails and weekly expense reports. What sets AutoExpenseFlow apart is its enterprise-ready architecture featuring human-in-the-loop escalation for exceptions, real-time policy enforcement, duplicate detection, and multi-currency handling. It reduces processing time from 15 minutes per invoice to under 30 seconds while maintaining 99%+ accuracy. The solution is particularly valuable for organizations struggling with compliance risks, audit preparation challenges, and rising operational costs in finance departments. Built with scalability and security as core principles, AutoExpenseFlow demonstrates how generative AI can transform back-office operations while providing immediate ROI through labor reduction, improved compliance, and real-time spending visibility."
23 Nov 2025