Getting things across the line.

Designing and shipping spec-driven, AI-assisted tooling. Most recent: a job-search agent that pulls postings straight from ATS boards, scores them with an LLM against a written profile, and emails a ranked daily digest.

• Custom adapters for common Applicant Tracking Systems, normalizing data into a common schema.
• Data ingested from desired companies based on a registry of companies and their corresponding Applicant Tracking Systems. Hash-based deduplication prevents disorganization and allows for idempotent database upserts.
• LLM scoring on fit and risk, data stored in SQLite, with a daily email digest using send-once tracking.
• github.com/ctwarrick/job-agent

Steered a legacy-to-migrated messaging middleware team through return-to-office and the aftermath of a major cyberattack while holding a reliable release cadence. Onboarded six developers and QA testers. Then stood up a new COBOL-to-Databricks ETL team.

• While coaching the middleware team, built a Flask/SQLite dashboard aggregating status outputs from 150+ production and pre-production servers after a Chef failure silently pushed outdated application versions to production. For on-call team members, this replaced a server-by-server hunt through status pages with one-click visibility into what was running where, enabling far faster incident triage and decreasing mean time to recover.
• Developed Python automation against the Azure DevOps REST API to generate and process QA approval documents and change orders for the middleware team's four separate applications, removing a manual gate that had helped lock the team into phase-gated, end-of-iteration deployments and clearing the path toward true CI/CD.
• After seven years with the middleware team, transitioned to stand up a new team writing ETL pipelines for a COBOL-to-Databricks medallion architecture. Persuaded senior leadership to backfill the previous role at full seniority rather than down-level it. Because I personally intervened to create a wider and more seasoned candidate pool and personally shortlisted candidates for management, we landed a former Senior Director with 20+ years of experience.
• While coaching the ETL team, used spec-driven development and AI to build a Flask dashboard that ingested YAML field-mapping files from the ETL pipeline and computed per-table Bronze-to-Silver mapping progress, giving leadership a real answer to “how far along are we?” that could be updated daily or with every YAML change.
• Created a Flask application to create cycle-time scatterplots and swarmplots from the Azure DevOps REST API, as well as Monte Carlo simulations for probabilistic delivery forecasting, replacing story-point guesswork with data-driven forecasting. Extended it with XmR process-behavior charts (statistical process control) for Work Item Age, Throughput, and Cycle Time, separating signal from noise and enabling data-driven conversations on team process stability. Introduced Monte Carlo and probabilistic forecasting to the Global Technology department in a 45-minute talk to 200+ attendees at exp.oCon, the annual innovation fair.
• Via AI-assisted development, migrated the Monte Carlo and ETL mapping Flask applications from on-prem VMs to Azure Container Apps behind an Application Gateway, with corporate vanity URLs for ease of use by nontechnical employees. Built GitHub Actions CI/CD workflows (among the company's first) to build Docker images, push to a Container Registry, and auto-deploy on approved merges to main. Coordinated subnets, VNet peering, and DNS with Datacenters and Networking teams and added health-check endpoints to eliminate gateway 502s.
• Co-authored formal Scrum Master performance standards adopted department-wide, improving evaluation and compensation for the entire role family, as well as an "Expeditors Lean-Agile Foundations" training curriculum used to onboard and upskill technology staff.

Coached four teams (test automation, legacy-migration middleware, reference-data ETL, freight accounting) through the company's SAFe adoption and the 2020 shift to fully remote.

• Took the middleware team from manual QA processes and ad-hoc releases to a reliable biweekly release cadence sustained for seven straight years, directly enabling 330+ operations branches worldwide and the company's strategic migration off legacy software.
• Built a Python bot for recurring Azure DevOps admin; forked by teams in Europe into an ongoing automation effort
• Led delivery of a major initiative automating worldwide freight delivery timetable updates and improving their accuracy; coached business stakeholders up to the Vice President level on SAFe to keep priorities and expectations aligned.
• Onboarded three QA testers with their primary Lean-Agile training. Mentored a career-changer into her first Scrum Master role through six months of direct shadowing; she's now a Release Train Engineer.