We have led high-tech delivery at every size — a 160-developer program across five onshore and offshore teams, mid-size squads of 50 to 60, tight teams of 8 to 9, and today a one-person, AI-augmented build shipping what used to take a department. The constant across all of it: clear architecture, people who know what to do next, and the discipline to lead with the written word so nothing gets lost in translation.
The founder is an Eagle Scout — Scouting’s highest rank, earned by a small fraction of those who start. It is where the leadership instinct was first formed: take responsibility, organize people toward a goal, finish what you start, and leave the campsite better than you found it. Decades later that is still how we run an engagement — we lead people on nearly every project, and we lead them on purpose.
From a program the size of a company department down to a single architect — the same standards at every scale.
Five ways of leading, matched to the size and shape of the team.
Running multi-team programs across time zones — up to five teams, two onshore and three offshore. Sprint planning and grooming structured so design, implementation, and testing each have a home, and so the data-science side and the development side actually understand each other.
For offshore teams with language gaps, we lead in writing — detailed emails, written designs, and documented decisions. It feels slower in the moment; in practice it is faster, because nothing is lost to a misheard call and everyone has the same source of truth to work from.
We lead from inside the code — primary PR approver, reviewer of full source bases, author of the sequence and architecture diagrams the team works from. Leadership that has read the code it is directing, not just the status report.
Solutions and business architecture that gives a team a spine to build against, plus the mentoring that raises the people around it — recommending architects on both sides, smoothing process, and making testable interfaces the default.
The newest scale: one architect leading a fleet of AI agents. With Claude, Copilot, and the agent SDK, a single person now ships what once needed a squad — a full SaaS platform, front to back, in months. Leadership of machines is still leadership.
Trunk-based development, comment and review standards, unit and integration testing strategy, and the candor to tell management when Kubernetes earns its keep and when it does not. Finishing what we start — an Eagle Scout habit that never left.
Two patterns we lean on — the full delivery spectrum, and written-word leadership across time zones.
The same engineering standards hold whether we are directing a 160-person program or shipping solo with AI. What changes is the leverage: more people, then better architecture, then better tooling — and today, a single architect orchestrating AI agents to cover the ground a department used to.
flowchart LR
P160["160 developers -- 5 teams, onshore and offshore"] --> P60["60 developers -- banking program"]
P60 --> P50["50 developers -- insurance GIS"]
P50 --> P9["8 to 9 -- streaming and edge squads"]
P9 --> P1["1 architect + AI agents -- full SaaS solo"]
P1 --> STD["Same standards -- architecture, testing, finish what you start"]
With offshore teams and language gaps, the spoken call is the lossy channel. We push decisions into writing — a clear email or design doc becomes the single source of truth that an offshore team can read, re-read, and act on without ambiguity. Slower to write, far faster to execute.
flowchart TD
DEC["Decision or design"] --> WRITE["Written down -- detailed email or design doc"]
WRITE --> SOT["Single source of truth"]
SOT --> ON["Onshore team"]
SOT --> OFF["Offshore team -- reads and re-reads, no language loss"]
ON --> EXEC["Aligned execution"]
OFF --> EXEC
EXEC --> FAST["Slower to write, faster to deliver"]
Five engagements, five team sizes, one set of standards.
Program lead over five teams — two onshore, three offshore — building a cardiac-device big-data platform on Hadoop. Organized sprint planning across time zones and kept the data-science and development sides translating cleanly to each other.
Business Solutions Architect and top technologist on a 60+ person mainframe-to-Azure program — owning the service topology and security architecture for a 10M-user system and setting the reference patterns the teams built against.
Solutions Architect leading a largely offshore team of 50 on a property-risk GIS platform — primary PR approver on the engineering database, author of the C4 and sequence diagrams, and the bridge between vendor, infosec, and the build teams.
Application Architect leading 8 to re-architect a national streaming provider, and scrum master over 9 building Go edge services for oil-and-gas field automation — small teams, high trust, fast cadence.
The new scale of leadership: a full AI-powered SaaS platform — Blazor web, .NET MAUI mobile, and a .NET backend orchestrating multiple Claude models — designed and shipped by one architect leading AI agents, the work a whole team used to carry.
Read the full case studyFrom 160-developer programs across onshore and offshore teams to a single architect shipping with AI, Leopard Data leads delivery at any scale. Corp-to-Corp engagements out of Plano, TX.