Resumé

View as PDF

James Nicholson's CV

Recent Highlights

Winner of the Llama Lounge Agent Hackathon — Built an agent-to-agent negotiation platform using Snowflake (data), CrewAI (agents), Composio (skills), and Skyfire (A2A payment) (Feb 2026)

Winner of the Qualcomm/Meta ExecuTorch Edge AI Hackathon — Built a mobile app with on-device Meta Llama 1B running on Qualcomm Snapdragon 8 Elite for voice-guided lab protocols (Oct 2025)

Experience

Harvard Chan School of Public Health

  • Sept 2019 – present

Lead Software Engineer

  • Aug 2026 – present
  • Served as technical lead on a multi-year platform redesign with two outside vendors, leading 12+ stakeholders through a 200+ test-case UAT process that surfaced 100+ bugs before launch.
  • Pushed early for a three-month launch delay on the redesign and secured buy-in from the project team; the delay proved critical to a stable launch.
  • Designed and built a fine-grained authorization layer on HarvardKey, Grouper, and SAML2, letting the school repurpose its existing platform as an intranet instead of buying a commercial solution.
  • Partnered with Harvard's identity and access management team to design a drop-in, backwards-compatible Okta identity integration, now the reusable migration path for school sites.
  • Took on operational ownership as the team shrank from three developers to one: became support lead for three backend applications, security and vulnerability lead, and owner of AWS patching and updates.
  • Initiated and led the migration of three backend applications off a shared local environment and into isolated ones, removing a coupling that had blocked updates across all three.
  • Authored and presented a next-generation WordPress tooling recommendation covering dependency management, automated testing, and monitoring, setting the standardized-workflow goal for the team's platforms.
  • Led the most complex DNS and hosting cutovers in a 2025 program that migrated dozens of websites, coordinating end users, Communications, and identity management to run migrations in parallel on schedule.
  • Mentored engineers through pair programming, code review, and design partnership, bringing two developers to productive contribution on the redesign and supporting a peer leading a separate site launch.
  • Led the IT department's adoption of AI-assisted development with a department-wide knowledge transfer; the team now uses AI for feature work, debugging, and browser-based testing.

Senior Software Engineer

  • Sept 2019 – Aug 2026
  • Built a nightly data integration pipeline pulling faculty profile data from multiple Harvard REST APIs into a relational database, serving as the authoritative source for 1,500+ researcher profiles on the school's public website.
  • Built a contact synchronization pipeline integrating the People Data Service REST API with an email marketing platform, automating nightly list management for faculty, staff, and students across multiple organizational sub-accounts.
  • Built cron-driven background automation jobs for user management and platform maintenance, including a process that verified and removed 1,000+ unauthorized accounts against an external identity API.
  • Improved production observability by configuring Sentry error monitoring, reducing accepted error volume by ~75% through event classification and noise suppression.
  • Completed a Master's Degree in Software Engineering from Harvard while working full-time.

Stanford Graduate School of Business

  • Sept 2015 – Sept 2019

IT Systems Analyst

  • Sept 2015 – Sept 2019
  • Developed and maintained a service request platform across multiple backend iterations, with automated scheduling, metadata tracking, and API integration with third-party class capture systems, saving hundreds of hours annually.
  • Built Python automation scripts that streamlined data management workflows and improved system reliability for technical and non-technical users.
  • Created Tableau dashboards from service platform data, enabling management to track usage metrics and support data-driven operational decisions.
  • Provided real-time technical support for high-stakes live events including MBA Admissions webinars and global distance learning programs serving 400+ concurrent attendees.

Education

University of Colorado Boulder, BS in Journalism

Harvard Extension School, MS in Software Engineering

  • Capstone Project: LLM-powered document extraction for Pfizer meta-analysis https://nicholsonjf.com/hectre

Skills

  • Languages: Python, SQL, TypeScript, JavaScript, PHP, Scala, C
  • Frontend: React, Next.js (App Router), Tailwind, React Native
  • Backend: FastAPI, Node.js, Express, REST APIs, Postgres, MySQL, Redis
  • AI: Anthropic & OpenAI APIs, Claude Code, MCP, multi-agent systems, LM Studio, ExecuTorch, local & edge inference
  • Infra: AWS, Docker, GitHub Actions, CI/CD, Sentry, SAML2, OAuth2 / OIDC, Okta

Projects

Information Extraction via Large Language Model (HECTRE)

  • Developed a web and CLI application leveraging LLMs and parallel processing to extract structured clinical trial data from PDFs, reducing manual processing from months to minutes.
  • https://nicholsonjf.com/hectre/

Multi-Class Image Classification: Transfer Learning vs. Standalone CNNs

  • Compared transfer learning against standalone CNNs for multi-class classification using deep learning techniques; implemented with Python, TensorFlow/Keras, and pandas.
  • https://nicholsonjf.com/transfer-learning/

Predicting Flight Delays

Using Akka Streams to Process MBTA Predictions

  • Built a Scala/Akka Streams pipeline consuming real-time transit data, demonstrating reactive programming, backpressure handling, and stream processing for big data applications.
  • https://nicholsonjf.com/mbta-arrivals/

Implementation and Comparison of Navigation Algorithms

  • Implemented and benchmarked A*, Dijkstra's, and other pathfinding algorithms for robotics and autonomous vehicle contexts, analyzing algorithmic complexity and performance tradeoffs.
  • https://nicholsonjf.com/navigation-algorithms/