I turn ambiguous operational problems into clear product decisions: understand the real workflow, isolate the riskiest assumption, choose the smallest useful release, and define the evidence that would justify scaling it. My computer-engineering background helps me collaborate closely with design and engineering, especially on AI-assisted workflows, data products, and internal tools where trust, reliability, and human judgment matter.
- Start with the user's decision, current workaround, and cost of failure.
- Separate observed evidence, assumptions, constraints, and open questions.
- Compare alternatives and make the tradeoffs explicit.
- Define outcome metrics and quality guardrails before building.
- Label results honestly as measured, simulated, or proposed.
- Treat limitations, exception states, and human handoffs as product requirements.
- AI-assisted B2B workflows with meaningful human review and traceable evidence.
- Data products that turn messy inputs into operational decisions.
- Internal platforms where reliability, permissions, and auditability are part of the user experience.
| Case study | Product question | What it demonstrates |
|---|---|---|
| Axiom Bee AI Inspection | How can AI accelerate industrial inspection without hiding uncertainty or removing human accountability? | Workflow framing, human review, evidence traceability, risk, and rollout thinking |
| Bakery Inventory Decision Support | How should a bakery prioritize inventory when demand and operating assumptions are uncertain? | Prioritization, service/cost tradeoffs, transparent assumptions, and decision support |
| Retail Analytics Decision Support | Which data and metrics help retail operators make better daily decisions? | Metric definition, data quality, analytical modeling, and dashboard design |
| DataHive Event Platform | What must a distributed system record so teams can trust and operate it? | Event contracts, auditability, reliability, and failure-aware platform thinking |
These repositories include prototypes and academic projects. Each case study distinguishes verified behavior from assumptions, synthetic inputs, and proposed improvements. I do not present simulated results as production outcomes.
Computer engineering, Python, JavaScript, SQL, cloud systems, analytics pipelines, computer vision, and UX-focused implementation. I use this foundation to ask better product questions and work effectively across product, design, data, and engineering.