I help teams ship production systems reliably — from inference APIs and event pipelines to embedded Linux and legacy modernization. Staff-level engineer, M.S. Physics, 15+ years across backend, embedded, RF, and safety-critical software. Two issued U.S. patents.
Problem: Manual review of incoming data was costing ~$1M/year in labor and missed issues.
Solution: Built Kafka + FastAPI pipeline routing events to ML models for automated classification and anomaly detection.
~$1M/yr cost reduction · Processing thousands of events daily · 50% faster turnaround
Problem: Manual API gateway configuration was slow, error-prone, and difficult to audit.
Solution: Architected configuration-as-code deployment of Tyk gateway with CI/CD testing to safely expose APIs while tracking changes.
Fast & safe API exposure · Version-controlled config · Auditable changes
Problem: Mobile field systems couldn't integrate with legacy SQL Server stored procedures. Development velocity stalled.
Solution: Replaced stored procedure logic with well-tested FastAPI services backed by Postgres, exposing clean REST interfaces.
Mobile-ready APIs · Testable service code · Unblocked mobile team
Problem: Quality issues were missed in manual inspections, causing expensive arbitrations and customer complaints.
Solution: Deployed OpenCV-based vision system and sensor fusion captures feeding ML models for real-time defect detection.
Early detection · Reduced arbitration costs · Integrated with existing production line
Focused on backend systems for AI applications: building the services that expose ML models as APIs, integrate inference into event-driven systems, and deliver predictions at scale. A few representative outcomes from prior work across commercial AI, defense, R&D, and safety-critical systems.
Connect ML inference to your existing streaming infrastructure. Route events to models via Kafka, handle fan-out patterns, and manage throughput and failure modes so your AI actually scales.
Build and deploy production APIs and microservices using FastAPI, Django, REST, and gRPC. Backed by Postgres, MySQL, Redis — integrated with your existing systems.
Ship backend services reliably with Docker and Kubernetes. CI/CD pipelines that deliver reproducible builds and safe rollouts for ML-serving applications.
On-device computer vision with OpenCV, embedded data collection, and sensor fusion. Detect issues before they become costly — proven to save $1M/yr in production.
Migrate legacy SQL Server to modern APIs. Translate stored procedure logic into well-tested services that integrate cleanly with mobile and web clients.
Build automated test pipelines for backend, mobile, infrastructure, RF systems, and safety-critical hardware. Verify contractor deliverables and acceptance criteria.
Requirements documents, technical proposals, and deliverable specifications. Clear written communication is part of every engagement.
Lead and mentor backend and mobile engineers. Slot in as architect, scrum master, or senior IC — whatever shape the engagement needs.
Secure over-the-air update infrastructure for embedded devices. K8s-based deployment with API wrappers and artifact management for safe, auditable fleet updates.
Targeted engagements such as test automation, ML pipeline design review, infrastructure debugging, performance analysis, or a focused build.
Fixed-scope delivery against defined acceptance criteria — for example, a streaming inference service, a data migration, or a test harness.
Ongoing staff-level capacity for teams that require senior infrastructure and AI judgment without a full-time hire.
Need senior backend/AI architecture help? Rescuing a stalled deployment? Building production-grade inference systems? Share what you're building, where you could use help, and your timeline. Replies within one business day.