Edge, Realtime & AI SystemsDelivered, Not Staffed
I design and build Cloudflare-native, realtime and AI-driven platforms for teams that need results quickly — without hiring internally.
Outside IR35 · Ltd company · Outcome-based delivery · Remote
Try the demo
Realtime voice AI • Deepgram + Gradium
Click to start a voice conversation
Requires microphone access
What I help teams ship
I take on time-bound, clearly scoped engagements where internal teams lack specialist depth or speed.
Edge & Cloudflare Systems
- Cloudflare Workers & Durable Objects
- Realtime WebSocket infrastructure at the edge
- Auth, rate limiting, queues, fan-out architectures
- Performance, resilience & cost optimisation
Realtime & Streaming Platforms
- WebSocket-based realtime systems
- AI streaming (text, voice, video)
- Low-latency dashboards & collaboration tools
- Voice & video orchestration pipelines
AI & Agentic Systems
- Agentic workflows (tool-using agents)
- Streaming AI responses
- Voice agents & conversational pipelines
- Prototype to production handover
Go, HTMX & Desktop Applications
- Go backends & APIs
- HTMX-driven web apps
- Go + HTMX desktop applications
- Internal tools & operational systems
Core stack: Cloudflare, Go, TypeScript, Hono, HTMX, Drizzle, Tailwind, WebSockets, AI (agentic, voice, video)
Realtime voice architecture
Sub-100ms round-trip latency through edge-first streaming pipelines
Selected platforms I've built end-to-end
Layercode — Realtime Voice Platform
- Designed and built the realtime voice platform for launch
- Low-latency streaming architecture
- Production-ready delivery under time pressure
MedAscend — AI Healthcare Platform
- Entire original platform architecture & implementation
- AI-driven workflows
- Full-stack delivery
Callwell — AI Voice & Calling Platform
- Complete platform build
- Realtime voice systems
- AI-powered call handling
Continuata — Platform Consultancy
- Entire platform design & build
- Cloudflare-first architecture
- Realtime & edge-native systems
In each case, I owned architecture, implementation, and delivery — not just individual tickets.
Background
I have operated as an independent consultant for over a decade, delivering production systems across finance, government, healthcare, and SaaS. My work spans AI platforms, realtime systems, and performance-critical architectures — with a focus on shipping, de-risking delivery, and transferring ownership to client teams.
HTMX Core Team
Delivering websocket support for v4 with Carson Gross
25+ years delivery
Production systems across enterprise and public sector
PhD-level research
Applied mathematics for production AI systems
Full delivery track record and consulting history.
Download CV — Stu Kennedy (PDF)Engagement model
All work is delivered as scoped consultancy, not team augmentation.
- Defined deliverables & milestones
- Fixed-term or fixed-scope
- Async communication
- No BAU or line management
- Handover & documentation included
Typical engagements
- 10-30 days total effort
- Remote delivery
- 2-3 days/week if required
Commercials
- Day rates typically £700-£900
- Outside IR35
- Weekly or milestone invoicing
Scope
Define outcomes, constraints, and deliverables
Deliver
Architecture and implementation to milestones
Handover
Documentation and transfer to internal teams
Exit
Clean completion on agreed terms
If you're looking to embed a contractor into an existing team or roadmap, I'm probably not a fit — and that's intentional.
Let's talk (if this fits)
If you need:
- Realtime or streaming systems
- Cloudflare or edge expertise
- AI agents, voice or video platforms
- A senior engineer to deliver, not embed
I only take on a small number of scoped engagements at a time.
© 2025 Stu Kennedy