John Dahl

Backend engineer with 10+ years building concurrent distributed services and data pipelines in Go, Python, and Java. Works on systems where throughput, memory behavior, and query latency are the binding constraints — most recently a Go engine that parses 31M+ packets from capture files larger than available memory, under bounded concurrency with flat memory usage. Experienced operating containerized services in production on Docker and Kubernetes, including diagnosing concurrency failures and resource contention under load.

Completed an 18-month engagement with Apple in July 2026, delivering one of the first production enterprise Model Context Protocol (MCP) server deployments.

Backend / Applied AI Engineer, Independent Consulting at NetworkForensics.AI (Network Forensics LLC) (2026-01 – Present)

Core engine open source: github.com/JohnDahl2/networking_software (MCP/LLM analysis tier in a private repo)

• Built a network forensics engine in Go from scratch, using goroutines and channels to process capture files larger than available memory — the ceiling that causes standard packet analysis tools to fail.
• Designed a bounded-concurrency worker pool that parses 31M+ packets from multi-gigabyte PCAP files with flat memory usage and configurable CPU throttling.
• Built the storage layer on TimescaleDB: batched packet inserts via Postgres COPY into time-partitioned hypertables, sustaining sub-second analytical queries over tens of millions of rows.
• Built an analytics API with cursor-based pagination, stateful job tracking, real-time progress streaming, and SHA256-based packet deduplication.
• Built a companion live-capture daemon in Go (gopacket) running concurrent capture sessions managed over a Unix-socket CLI, streaming packets off the wire into the same analytical store.
• Packaged the Go services, live-capture daemon, and Python operator UI as a multi-container Docker Compose deployment, with a documented REST API and reproducible single-command setup.
• Decoupled the backend into an LLM-orchestrated analysis tier (Claude API), exposing the Go/TimescaleDB telemetry pipelines through MCP query tools with a sanitization layer for regulated data.

Backend Software Engineer (Contract via Pinnacle Consulting) at Apple (2025-01 – 2026-07)

AI Data Services

• Designed and shipped a production Model Context Protocol (MCP) server layer in early 2025 (Python / FastMCP) — among the first enterprise MCP deployments — giving non-technical stakeholders natural-language access to operational data via concurrent backend lookups.
• Migrated reporting workflows from synchronous REST to an event-driven, LLM-orchestrated query architecture, converting recurring manual analyst requests into self-service natural-language queries.
• Built Python data ingestion pipelines and threshold-based alerting for data center operational metrics, tuning for memory footprint and low-latency threshold verification.
• Developed an automated monitoring engine for carbon usage and expenditure thresholds, replacing periodic manual review with real-time notification workflows.
• 18-month engagement concluded on schedule at project completion in July 2026.

Backend Software Engineer at Eagle Eye Networks (2023-05 – 2025-01)

• Diagnosed and fixed instability in a high-concurrency async Python consumer on Apache Pulsar, eliminating recurring service crashes and container restarts in production.
• Separated a co-located Django API and async message consumer into independent services, resolving the contention that had been degrading both.
• Deployed and operated containerized Python services on Docker and Kubernetes, diagnosing pod restart loops and resource contention in production.
• Refactored async Python to correctly handle concurrent database and API calls, stabilizing notification delivery across SMS, email, Slack, and ticketing integrations.
• Built Django REST APIs for a cloud video management platform serving 2,450 connected cameras across 60 enterprise clients.
• Extended public REST APIs for third-party and mobile clients, supporting remote viewing, motion detection alerts, and dynamic bandwidth management.

Backend Software Engineer (Contract via Insight Global) at Wells Fargo (2021-11 – 2023-05)

• Built a centralized Python/Django platform delivering the enterprise's first consolidated view of global financial risk positions, replacing fragmented siloed reporting.
• Ingested real-time market data streams to generate market, currency, bond, and commodity risk assessments used directly by the CFO's office.

Software Engineer at Network Data Science (2020-04 – 2021-10)

• Built Python machine learning models (scikit-learn, TensorFlow) to detect anomalous patterns in network telemetry, enabling proactive mitigation.
• Built Plotly Dash dashboards surfacing infrastructure performance trends directly to stakeholders.

Software Engineer at CBC Financial & Insurance Services (2018-06 – 2020-03)

• Automated multi-source data ingestion, cleaning, and transformation pipelines in Python, cutting report delivery cycles by 30%.

Software Engineer at Pioneer Analytics and Consulting (2016-01 – 2018-05)

• Built Java REST clients harvesting real-time telemetry from distributed internet agents into downstream performance monitoring and anomaly detection pipelines.
• Developed Spring Boot microservices for real-time data ingestion and external API integration, supporting 240 downstream consumer applications.

M.S. in Analytics – Wittenberg University

B.S. in Economics – The Ohio State University