What started as Java debugging in a college coffee shop became 6+ years shipping production systems. I build backend services in Python and Java, cloud infrastructure on AWS, and multi-model LLM architectures on Bedrock, with the guardrails, cost budgets, and observability that hold up under real traffic. My instinct is toward systems that operate themselves and AI that ships to real users, not demos.
LinkedIn Profile
Professional Experience
Two tenures across regulated workloads: SOX-compliant banking at JPMorgan Chase and HIPAA-compliant healthcare at Quality Health Care Group.
JPMorgan Chase & Co.
Software Engineer (Banking)
April 2024 - Present
Platform engineering across private and public cloud for SOX-regulated banking services. Jenkins, Spinnaker, Terraform, AWS (Lambda, ECS, Step Functions), Cloud Foundry.
- Brought a business-critical web application (9 components) into compliance with firm engineering standards and closed its monitoring gaps: liveness/readiness health checks, a post-deployment validation gate, Prometheus telemetry, Celery worker outage inference from RabbitMQ consumer counts, and per-container CPU alerting in Grafana across 9 production applications, surfacing a saturated ingestion container that host metrics had hidden.
- Built an MCP toolkit of 200+ Python tools that lets AI agents act on internal platforms through the firm's single sign-on flow, wrapping Jira, Confluence, Bitbucket, Grafana, ServiceNow, LDAP, Jenkins and Cloud Foundry, so agents query, operate and audit production systems without bespoke integrations.
- Architected an in-process AI application on three Bedrock models (Haiku/Sonnet/Opus) for pre-release automation planning, with model-version audit stamping and deterministic fallbacks, making a 55-step manual release checklist about 60-70% automatable.
- Standardized delivery across 9 Django applications: declarative Jenkins and Spinnaker pipeline templates with blue-green deployments and proven automated rollback for zero-downtime releases, plus a sandbox tier ahead of develop so database changes are proven before shared environments.
- Led a relational schema cut-over across legacy, API and frontend layers with an authoritative naming map, sequenced migrations and rollback-aware plans verified against production row counts.
Quality Health Care Group Inc.
Software Engineer (Healthcare)
June 2018 - April 2024
Sole-developer build of HIPAA-aligned healthcare platform. Django REST Framework + React/TypeScript → Cloud Build (19-step CI/CD) → GCP/AWS.
- Architected HIPAA-aligned healthcare platform end-to-end (Python REST API + React/TypeScript) and a 19-step Cloud Build CI/CD pipeline with automated rollback and distributed locking against concurrent deploys.
- Engineered JWT authentication with RBAC across 4 clinical role tiers, token rotation, and 7-year HIPAA audit trails.
Multi-Model Routing
Cheap models decide yes/no; hard questions climb a ladder. Production pattern from the Healthcare Compliance Platform: multi-model Bedrock routing cut monthly inference costs 78% on HIPAA-regulated document processing. Explore how different request types flow through the tier ladder.
Rates shown are illustrative for tier comparison, not published pricing. Current rates: aws.amazon.com/bedrock/pricing
Selected projects
Healthcare Compliance Platform: production GenAI on HIPAA infrastructure. ARC Modality: multi-tenant AI infrastructure built as independent R&D. Plus earlier social-media + cloud work at Xoobug.
Certifications & Credentials
Currently pursuing the AWS Certified Generative AI Developer Professional cert; target October 2026.
AWS Certified Generative AI Developer Professional
In Progress, target October 2026Exam code AIP-C01. Covers foundation models, multi-model orchestration, Bedrock integration patterns, prompt engineering, RAG architectures, responsible AI, and production GenAI deployment.
AWS exam details →