Agentic AI Development
Designed agentic workflows for autonomous decision-making and AI-powered process automation. Integrated conversational AI with custom LLM systems and built agentic automation frameworks with HIPAA/GDPR-compliant data pipelines.
Agentic AI is fundamentally different from traditional ML. Where predictive models answer 'what will happen?', agentic systems answer 'what should be done?' and then do it. I've built agents that monitor production systems, detect anomalies, and self-heal without human intervention. The reliability engineering required is closer to distributed systems design than data science.
At Notes AI, I designed a multi-agent architecture where specialized agents handled specific domains: a compliance agent, a data quality agent, and an orchestration agent that coordinated them. Each agent had a narrow responsibility, clear success criteria, and graceful failure modes. The hardest problem was coordination making sure agents didn't conflict when acting on the same data simultaneously.
Human-in-the-loop design is non-negotiable for high-stakes agentic workflows. I implemented confidence thresholds that determined whether agents acted autonomously or escalated decisions. Low-confidence situations surfaced to human reviewers with full context not just the decision, but the agent's reasoning chain. This transparency built trust that allowed gradual automation of higher-stakes decisions over time.
The practical impact was significant: tasks that previously required daily human review now ran fully autonomously, with humans reviewing exceptions rather than approving every action. This shifted the organizational model from process execution to exception management a fundamentally different and more scalable operating model.
At Notes AI, the multi-agent architecture I designed used specialised agents for compliance, data quality, and orchestration, each with narrow responsibility and graceful failure modes. The hardest problem was coordination, ensuring agents did not conflict when acting on the same data simultaneously, which I solved with distributed locking and event-sourced state tracking.
The HIPAA and GDPR-compliant data pipelines at Notes AI ensured that agentic workflows processing healthcare data met regulatory requirements for audit logging, data residency, and patient privacy. The confidence-threshold escalation system meant low-confidence decisions surfaced to human reviewers with the full reasoning chain, building the trust that allowed gradual automation of higher-stakes decisions over time.
Ecommerce & Digital Automation
Related Skills
Explore the Full Portfolio
This is the complete portfolio of Sufi Khan Sulaiman, a technology leader specialising in B2B commerce and digital automation. Start from the Home page for the overview, then move through two decades of career experience across FLIR Systems, Lorex Technology, and 1c Platform, and the full catalogue of project case studies spanning headless commerce migrations, AI recommendation engines, and multi-channel fulfilment systems.
The skills and certifications page maps the technical and leadership capabilities behind the work, while the articles and the knowledge base break down the thinking into actionable frameworks. For hands-on learning, the tutorials and applications sections cover practical builds from front-end fundamentals to full-stack web apps.
For consulting engagement, the expertise page outlines service offerings, the ecommerce hub covers platform architecture and automation strategy, and the ecommerce guide (PDF) is a downloadable 55-page field manual. When you are ready to talk, the contact page is the direct line.