Enterprise Agile Coaching | SEC
Conducted an SEC agile readiness survey, briefed the CIO on the recommendations, and authored the SEC agile playbook.
Agile Project Management (per se) was a small component of this enterprise-level engagement with a federal agency—ensuring the small team is performing according to contractual obligations.
For this engagement, I capitalized on my 20 years of experience in project management and seven years as an agilist to help an organization reenvision a new operating and governance model for managing IT projects, moving away from their traditional waterfall framework to a new agile paradigm.
The consulting engagement required communication (interviewing and influencing) at all levels of the organization - from rank-and-file PMs to the CIO. Written materials (presentations, meeting minutes, user stories, Confluence wiki pages, newsletter issues, and 80+ page agile playbook) were also shared with all levels of the organization, passing several levels of reviews and quality checks.
I coached the first formal agile project for building an online catalog system for storing, organizing, and searching metadata for large datasets. The team achieved a record-breaking success for delivering an operational MVP in three months after the start (compared to the 6-9 months for similar-size government initiatives).
Software Delivery Portfolio Management
Oversaw a portfolio of eight federal government programs, 160 personnel, and a $15.5-million yearly budget.
I fulfilled the P&L responsibilities for a portfolio of eight software delivery programs for my client who was contracted with the federal government.
My immediate team consisted of three program managers and six project managers. My organization totaled ~160 people consisting of software engineers and personnel with related skillsets.
All programs were at various levels of agile maturity (depending on the client's maturity and needs). My responsibilities included ensuring continuous learning and improvement of all programs.
Independent of the individual program execution, I chartered and led the formation of the corporate agile, DevOps, and cloud engineering center of excellence, where we developed prototypes and demo solutions.
Achievements in 2017:
• Developed a budget underrun of $1.7 million.
• Generated a net revenue of $1.6 million.
• Facilitated a 25% increase in CPAR (contractor performance assessment reporting) ratings.
• Earned numerous client kudos and social media shout-outs.
The Common Securitization Platform
Served as an agile/SAFe champion, DevOps lead/PO, an integration manager, and release manager.
The Common Securitization Platform is a new secondary mortgage market utility developed by a Fannie Mae-Freddie Mac joint venture (2012-2018).
I joined the team in 2012 as a 1-person scrum team (one microservice and an ESB) and left in 2016, (12 scrum teams, ten event-driven services, an ESB, three supporting unities, two major release branches, and fully automated DevOps processes for cloud deployment).
My role changed, evolved, and grew over four years.
Roles: Product Owner | Integration Lead | Platform Release Manager
Significant Achievements:
• Managed the MVP version of the RulesEngine by planning an XML extension of the MISMO data format and defining and building the initial set for 64 data validation rules.
• Defined the processes and SLAs for fulfilling ops requests, and automating any repeatable requests.
• Drove efficiency and process improvement supporting a team of ~200 developers/testers.
• Established new processes for PI planning, dependency analysis, impact analysis, and integration testing.
• Designed the release management processes, versioning and branching strategy, stakeholder communication, planning, and coordination.
• Championed SAFe adoption for scaling in an agile manner.
Corporate Functions | Production Support
Formed the team, defined the process and procedures, and eased the knowledge and operations transition to an MSP provider.
Role: Operations Management and Production Support
I managed the operations and provided support for production for Fannie Mae's corporate functions systems. These systems included one PeopleSoft HRMS instance (supporting all HR functions for ~6,000 employees) and five PeopleSoft Financials instances supporting various accounting and financial processes (including the remittance service for ~$20-billion monthly remittances of mortgage service payments).
I segregated and formed an ops team of 12 engineers. I also trained them in ITSM best practices that included developing, documenting, and implementing our standard operating procedures, and ultimately—package and hand-off everything to an external master service provider retained by Fannie (all the while, performing the daily ops duties).
I did not have a single severity-1 incident for approximately one year while in this role.
Impaired Loans Accounting System
Handled the greenfield implementation of a new accounting system for Fannie Mae in only six months.
The System Implementation Tasks:
• Integrated and customized a large COTS accounting package.
• Built pre- and post-processing ETLs.
• Recreated all 20+ interfaces.
• Rebuilt 30+ reports for a 30-person team.
The project had a financial impact of $1.2 billion in impaired loan booking.
Fannie Mae GAGO Restatement
Cured the riskiest project of a $2-billion Fannie Mae restatement program.
The "project" was a series of monthly releases for fine-tuning and rerunning the financial predictive models used to calculate the fair value of a Fannie Mae book of business. ( I consider it my first experience PMing an AI/ML project.)
The high complexity of the financial models and algorithms, the trial and error nature of the model evaluation, and the tuning process created a communication and governance gap between the FNMA PMO and the small business unit of financial engineers developing the models.
Balancing interpersonal and technical (math) skills, I was able to close the gap and bring the workstream into compliance with all new SoX requirements and PMO target milestone schedules.
MLOps Prototypes
Coached and led four teams in incorporating agile and DevOps techniques for solving 4 different AI/ML problems.
Each of the projects was a 2-3 week hackathon-type endeavor for developing a small MVP/POC. For each one, we were able to integrate AWS SageMaker notebook for batch-processing predictions, delivering data to a PostgreSQL/Elasticsearch datastore and microservice architecture application deployed on AWS OpenShift clusters, through fully automated CI/CD pipelines.
Two out of the four solutions required the integration of an ensemble of algorithms for NLP, entity resolution, and photo recognition. (The third was an NLP and unsupervised clustering of corpus data.)
Granted, each solution was small and somewhat synthetic, my teams and I demonstrated a full command of the technologies and processes for delivering ML-enabled solutions in a very short time period.
Miscellaneous Projects
Developed ~12 small products from scratch for various clients of a small software development services company.
I served as the IT director of a small systems integrator, personally overseeing all projects from proposal response, discovery, and scoping through final client acceptance.
During my tenure, my team developed ~12 different solutions with various levels of complexity. Granted, all of the projects were relatively small and waterfall and developed in technologies from the late 90s/early 2000s. However, I still value this experience tremendously because it allowed me to experience and learn full product development and project management lifecycles—something that very few PMs can claim.