Darren Hagman

Managed several teams with over 25 developers, testers, and DevOps staff in developing the initial MVP offering of our ParaTransit solution containing server-based booking, scheduling, dispatch, and vehicle management modules for scheduling and managing customer trips, a mobile driver application and hardware for executing routes, and a mobile rider application for booking trips

Managed several teams in developing and deploying the AVLM portion of a customized solution for the New York City Transit Authority (NYCTA) to support their Access-a-Ride service. The solution will be deployed to over 1500 vehicles and is currently deployed to over 400 vehicles.

Technologies used include AWS, PostgreSQL, DynamoDB, Node.js, React Native, Lambda functions, Kubernetes, Kafka, Ember, Android OS 8.0, custom Android tablets

• 10/10 — Seeded Regressions Caught in Mutation Testing
• 4 — Source Documents Under Automated Claim Control

Producing tailored senior-level application packages at volume meant every metric, title, and date had to stay accurate and traceable to evidence. Manual drafting was slow and allowed unverified or inflated statements to drift into documents. The real problem was quality control at scale, not writing speed.

I built a controlled document-generation pipeline using ChatGPT, Claude Code, and Python. A canonical claim-controls file defines approved wording and forbidden phrasings; a separate employment-history file anchors every date. A Python audit suite checks each generated package against those controls and fails closed on any violation, with regression tests covering the checks themselves. An independent adversarial review pass scores each package before submission. Independent applied work, not a production client deployment.

Every metric in a generated package is now traceable to an approved source, and the audit has caught claim drift that manual review missed. Controls and their regression tests run on every package before submission.

Managed two teams with a total of 15+ developers, testers, and business analysts to deliver the initial launch and numerous major releases of the Staples Copy & Print website (http://design.staples.com) which replaced the VistaPrint site and increased sales and revenue.

Managed a team of 15 + developers, testers, and DevOps in delivering the initial and subsequent releases of the write once read many (WORM) server application which exports archived customer messages from the Archive 10 system and makes encrypted tape backups of the archived data on a Quantum Tape Archive system for offsite storage.

Technologies used include Java, Thrift, Ansible, Jenkins, Zabbix, and Selenium.

Drove pre-project consulting with a senior architect to estimate and plan the retirement of the CIP system, which represented a downtime risk of €30 million if the system crashed and couldn’t be recovered. This lead to the approval of the project.

Led a team of 14 locally and remotely based IT specialists in successfully retiring the Distribution List Management tool.

Led a remote team in retiring 14 ( out of 23) workflows. Managed budgets (€1.2 million/year) and timelines.

As part of Boeing’s GoldCare Cloud-based SaaS offering, managed application development of several products.

Built and mentored strong project teams while establishing strengthened relationships with clients, resulting in increased business revenue (doubled purchase order dollar amount yearly) and first-class customer satisfaction ratings. Managed yearly budgets reaching $4 million per year.

Managed multiple releases of Situational Awareness which is a web-based dashboard that displays real-time flight and maintenance event data to enable operations and planning staff to predict and deal with maintenance and flight issues.

Managed several releases of Performance Metrics, which is a web-based business intelligence application that hosted dashboards and detailed reports relating to flight performance, GoldCare and MRO maintenance performance, and parts usage and reliability.

• 10 — Program Increments Organized for Enterprise SCADA
• 32 — Releases Supported Across the Trains

The Enterprise SCADA release train needed its leadership’s confidence rebuilt. A director was covering the release train engineer role after the prior arrangement failed, and senior leaders were attending routine working meetings because planning, dependency, and risk information could not yet be trusted from the train itself.

Took over the Enterprise SCADA train as release train manager. Led PI-planning preparation and execution, introduced meeting structures and dashboards, and managed dependencies, risks, and readiness as standing disciplines rather than escalations. Across the trains supported, I helped coordinate releases with development managers and product stakeholders, and played a substantial role in carrying releases through their closing QA audits. Enterprise SCADA was organized through 10 program increments.

I earned stakeholder confidence: directors and senior VPs stopped attending routine meetings once trust was restored. I supported 32 releases across six product areas as the combined output of the trains supported.

• 1 to 3 — Teams After Replanning and Escalation
• 12 Months — Engineering Initiative Recovered and Relaunched

The AIM engineering project had been rejected as a formal program in favor of a backlog-mode approach, with one team working on features continuously. It never gained traction, and when the person handling it retired, the work was handed over, stalled, with no plan showing whether the committed scope was reachable.

Reorganized backlog grooming and prioritization, then built a realistic timeline from the existing approach and actual capacity. The resulting completion date was unacceptable to stakeholders, which was the point: it exposed the gap between expectations and resourcing. Used that evidence to escalate resource needs and help move the work under engineering as a reorganized program, after which two further delivery teams and an engineering-side development manager were added.

The effort expanded from one team to three, produced an acceptable plan, and was relaunched under experienced engineering management, then handed to a Europe-based program manager for continued delivery.

• 4-5x — Revised Project Estimate vs. Original
• 2 — Delivery Teams After Replanning

Quick Quote sat within the $4+ million Dealer Services Automation Program, giving vehicle dealerships across Canada quotes on replacement, creditor, and extended warranty insurance. The project had been significantly underestimated, and the external delivery team was too junior for the work in front of it.

Validated the skill gaps with the architect rather than treating the estimate as merely optimistic. Secured a more senior developer and a full-time vendor architect, then re-estimated the work with the team and the product owner. Obtained additional budget on the strength of that re-estimate and added a second delivery team. Recovery rested on the replanning and the two teams together, not on the staffing change alone.

I delivered Quick Quote with an improved user workflow and improved quote-estimation logic. The revised project estimate was four to five times the original, and the replan improved velocity and set a realistic schedule and budget.