Intelligent systems that do the work your business doesn't have time for.
I design and build AI agents, ML pipelines, and automation that turn repetitive operations into reliable, measurable outcomes — engineered end to end.
What I build
Four disciplines, one engineer — so the system that ships is coherent, not a patchwork of vendors.
AI Agents
Autonomous and human-in-the-loop agents that reason over your systems — triaging, drafting, deciding, and escalating with full auditability.
- Tool-using agents
- RAG & retrieval
- Guardrails & evals
ML Pipelines
Production-grade pipelines for training, evaluation, and inference — built to be observable, versioned, and cheap to run.
- Feature & data pipelines
- Model evaluation
- Deployment & monitoring
Automation
Workflow automation that removes the manual glue between your tools, so your team spends time on judgment, not data entry.
- Process mapping
- Integrations & APIs
- Error handling at scale
Data & BI
Clean data models and dashboards that give you a single, trustworthy view of what's actually happening in the business.
- Warehouse design
- ETL/ELT
- Dashboards & reporting
Systems I've designed and built
A selection of the agents, pipelines, and ML systems I've built end to end — two of them running live and available to walk through on a call.
How an engagement runs
Discover
A short, focused engagement to map the workflow, the data, and the real constraint — not the one everyone assumes.
Design
A concrete technical plan: what's an agent, what's a pipeline, what's out of scope for now. No architecture for its own sake.
Build
Weekly shippable increments, not a black box for eight weeks. You see working software early and often.
Operate
Monitoring, evals, and documentation handed off cleanly — so the system holds up after I'm no longer the only one who understands it.
What clients say
“Suny didn't just build what we asked for — he pushed back on scope twice because a simpler version would ship faster and hold up better. That judgment is rare.”
“We'd tried two other contractors before Haqbil. The difference was the pipeline actually worked in production, on day one, without a month of firefighting after handoff.”
“Fast, direct, and genuinely good at explaining trade-offs to a non-technical founder. I always knew what I was paying for and why.”
Have a workflow worth automating?
Let's spend thirty minutes figuring out whether an agent, a pipeline, or something simpler is the right fit.