I am Rume
Dominic.
AI engineer and builder.
I build AI agents you can check and trust, and I take them all the way from a first prototype to a system that runs in the real world.
AI engineer with a track record of taking AI prototypes and turning them into reliable, cloud-hosted systems, across financial automation, subscriber analytics, and SaaS.
I do not just write code. I design, build and ship the whole thing. I build LLM, RAG and agent systems that take over document-heavy work, help people make faster calls, and turn a design into real money saved or earned.
My work on verifiable agent accountability placed third at the 2026 Build on Canton hackathon in London, judged by a panel that included the Canton Foundation's founding chairman, and the tool is listed in the community Canton Developer Hub (the canton-network-devs catalogue).
LLM & Agentic Systems
Orchestrating multi-agent workflows, tool usage, and prompt pipelines for autonomous execution and scale.
Machine Learning (RAG)
Building high-precision retrieval systems, OCR data extraction, and semantic search architectures.
Verification & Blockchain
Implementing cryptographic hashing (SHA-256) for tamper-evident documentation and IP provenance.
Product Finalization
Shipping dockerized SaaS that scales, on Google Cloud Run, Vercel and Next.js.
Know Your AgenticAI
Spending mandates enforced inside a Daml choice body on Canton, so the control cannot be bypassed by the service calling it, and every attempt, refusals included, becomes a hash-chained receipt an auditor can verify. Listed in the community Canton Developer Hub (the canton-network-devs catalogue), with three pull requests merged by the hub maintainer, and third place at the 2026 Build on Canton hackathon.
Local 7B LLM Email Processor
A personal passion project running a local 7B model via Ollama. It automatically processes my inbox, deletes spam, leaves normal messages unread, and flags receipts for Paperless-ngx.
VeritaPort (Layer-3 Verification)
Designed a cryptographic verification layer for tamper-evident document integrity, combining SHA-256 hash-chain verification and product-level tamper detection.
Matty's Place (Tenant Hub)
Developed an automated property management and tenancy tracking system tailored for Houses in Multiple Occupation (HMO).
Eagles Den Analytics
Built a five-model LLM analytics workflow for subscriber decision-making, including market-signal generation, Paystack billing, Telegram / WhatsApp delivery, and Cloud Run deployment.
AI Code Quality Auditor
Built a Python evaluation instrument measuring where AI coding agents fail across hallucination, code complexity, duplication, and security-density indicators, with 600 controlled measurements.
Microsoft Azure AI Apps & Agents Developer
AI Governance & Responsible AI
Python (University of Michigan)
UK Patent Application GB2611754.9 (filed 2026, pending), the KYA Framework for verifiable AI accountability and provenance.
Third Place, Build on Canton Hackathon
Placed third, as Vorem, at the largest in-person Canton Network hackathon of the year, nearly 100 participants across more than 20 teams, judged by a panel that included the Canton Foundation's founding chairman. Published in Cantor8's official recap.
Agentic AI Masterclass
Led an Agentic AI masterclass as the main speaker, for students, business owners and professionals.
Frontier Tech X: Agentic AI
Invited speaker at the prestigious University of Birmingham "Old Joe" for Frontier Tech X, discussing the horizon of Agentic AI and blockchain verification layers.
AI Dominance Event
Hosted and spoke at a large event showing how to use AI to get ahead, with a big in-person turnout and real case studies.
Creator Economy & AI Storytelling
Worked with Jali and Empower to teach creators how to use AI in storytelling and media work.