Jonathan Camp

Platform engineering for healthcare technology under FedRAMP Moderate and FIPS compliance requirements, from cloud infrastructure to data pipelines and internal tooling.

• Cut event volume 95% with a real-time change data capture pipeline (Django signals, Kinesis Firehose, S3, Athena, dbt) using incremental MERGE on Iceberg tables, keeping load off production APIs.
• Deployed a FIPS-compliant analytics platform (Apache Superset) in AWS GovCloud with ECS Fargate behind load balancers in auto-scaling groups, encrypted Aurora PostgreSQL, automated secret rotation, custom SSO integration/user provisioning, and CloudWatch observability, all provisioned with Terraform.
• Built a proof-of-concept AI operations platform bridging Claude to enterprise monitoring via MCP on a serverless AWS stack (Lambda, AppSync, DynamoDB, Bedrock).
• Replaced manual workflows with production services built in Go, Pocketbase, and React; operate Keycloak SSO infrastructure with custom provisioning for downstream systems.

Cross-functional development across distributed cloud infrastructure, from reliability and cost work to customer-facing features.

• Led migration from a legacy Java monolith to microservices, with Go services speaking gRPC over AWS App Mesh and EventBridge, SNS, and SQS handling async work. Extracted services incrementally along API boundaries; cloud costs fell by over 50% by the time the monolith was retired.
• Maintained and optimized the widget-keyed Kinesis partitioning behind billions of daily content recommendations, keeping widget×geo×device aggregate updaters on bounded, ordered state per shard while raw events flowed intact into Snowflake; the aggregates fed advertiser dashboards, bid pacing, and accounting.
• Migrated the event pipeline from Kinesis to Kafka for cost and operational ownership, after managed-service failures left us with no remediation paths within budget. Moving consumers onto Airflow made the data flow visible, down to individual latencies, side effects, and where things stalled.
• Ran low-latency serving state on a replicated, sharded Aerospike cluster, with ElastiCache caching content-recommendation results, and tuned CloudFront CDN caching to cut data-transfer costs.
• Implemented distributed locking on the ad-delivery path with DynamoDB conditional writes, so concurrent spenders could not push campaigns past budget and a stale lock holder could not commit over newer state.
• Instrumented the microservices with Prometheus metrics, Grafana dashboards, and OpenTelemetry traces into Datadog, then used what the telemetry exposed to raise throughput and cut latency and error rates.
• Load-tested services with Gatling and failure-tested in production via live region shunting, rerouting traffic away from a degraded region and verifying the system held under the shifted load.
• Converted legacy infrastructure to infrastructure as code (Terraform, CloudFormation); modernized a legacy PHP UI with React microfrontends via Module Federation.

Security monitoring systems and incident response in a managed security services environment.

• Built out a distributed, event-driven security monitoring system ingesting reports and data into dashboards for analysts, working in Docker on Azure with ASP.NET Core, Angular, Solr, and SQL Server.
• Conducted forensic analysis, root-cause analysis, remediation, and reporting for client security breaches, malware attacks, and vulnerabilities; supported expert-witness testimony in intellectual property litigation.

Led a team of five building a serverless advertising automation platform handling $50k–$100k in daily ad spend.

• Re-architected the platform from an Apache Storm cluster to a serverless AWS stack (Lambda, API Gateway, Step Functions, DynamoDB), saving thousands monthly in operating costs.
• Implemented immutable infrastructure with CloudFormation, Secrets Manager, and CircleCI deployment automation, enabling identical staging, QA, and production environments.
• Provided 24/7 operational support with root-cause analysis and systematic prevention of recurring issues.

Financial analysis platform built on a domain-specific language for quantitative research, using Java, Groovy, Spring, RESTful APIs, and automated testing.

• Replaced an outsourced PDF report-generation system with an in-house Java API, Angular frontend, and D3 visualizations, saving tens of thousands annually with page-for-page parity.
• Championed Dynatrace APM across a distributed architecture after identifying observability gaps, building the proof of concept that won buy-in and training the team on investigative workflows.
• Collaborated with quantitative analysts to implement mathematical models using appropriate software design patterns.

Agile development on medical automation systems in C#, ASP.NET, VB.NET, and Java.

• Developed services on a team building a medication delivery network platform (ASP.NET MVC, service-oriented architecture, domain-driven design), with testing from unit through performance and end-to-end.
• As Product Support Specialist II (2008–2011), diagnosed and resolved complex production issues across multiple languages and paradigms through root-cause analysis and reproduction test cases.