When a system is too slow, too tangled, or too important to guess at.
Most software teams eventually find themselves with a system they can no longer fully reason about. A database that has grown slow in ways no one has had time to trace. A service that fails intermittently, for reasons that don't survive the first few obvious theories. A codebase that has outlived the team's complete understanding of it. The work usually still gets done — but it gets harder, slower, and riskier, and the cost of a wrong guess keeps rising.
That is the work I do: diagnosing large, complex software systems, and figuring out what is actually wrong before anyone commits time and money to a fix.
I'm David Berube. I've spent over twenty years building, scaling, and troubleshooting software systems, including as co-founder and technical lead of Casting Frontier, a casting platform whose software was used across much of the Fortune 100 before its acquisition in 2020. I work with companies around Massachusetts and New Hampshire, and remotely where it makes sense.
What I help with
I focus on the problems that reward experience and judgment rather than raw effort — the ones where understanding the system correctly is most of the battle.
Database performance
Queries that have grown slow as data and load increased, indexing and schema decisions that no longer fit how the application is used, and the gap between "it worked fine at launch" and "it's a problem now." This has been a through-line of my work for two decades.
Complex system troubleshooting
Production problems that resist the obvious explanations — intermittent failures, performance that degrades under conditions no one has pinned down, behavior that doesn't match anyone's mental model of the system. Finding the real cause is often the hard part; the fix tends to follow from it.
Understanding inherited and overgrown codebases
Systems that have outgrown the team's full comprehension of them, whether through turnover, age, or organic growth. I'm unusually good at coming into a large, unfamiliar codebase and developing an accurate picture of how it actually works — not how it was meant to work.
Technical strategy and architecture review
Decisions that are expensive to get wrong and difficult to reverse: whether to rebuild or repair, how to plan a migration off a system you've outgrown, where the real risks sit in an architecture, and what a sound path forward looks like. This includes software business strategy, where the technical and business questions are tangled together.
How engagements work
Most engagements start the same way: a conversation about the system and what's going wrong, followed by a focused assessment. The goal of that assessment is a clear, honest picture of the problem and a recommended path forward — including, where it's the right answer, the option that costs you the least.
I try to be direct about trade-offs rather than selling a predetermined solution. Sometimes the right recommendation is a substantial piece of work; sometimes it's a small change that resolves most of the pain; sometimes it's that the system is fine and the problem is elsewhere. You're paying for an accurate answer, not a foregone conclusion.
When a diagnosis leads to implementation work, that building is handled through Durable Programming, a separate firm where I work with a small team of full-stack engineers. That keeps the advisory work and the implementation work cleanly separated — you can engage me purely for diagnosis and strategy, take the plan to your own team, or have Durable carry it out. The choice is yours, and there's no assumption that an assessment turns into a larger project.
What Clients Say
David Berube is highly recommended... He single handedly saved our project from imminent ruin.
—Joey Rubenstein, President, Casting Frontier
David is an unstoppable force. He can navigate large unfamiliar codebases and untangle ways to maintain and modernize them in a way that I have rarely seen matched... Highly recommended!
— Elliot Murphy, VP of Engineering, Posit PBC
Background
I began working as a software developer in 2001 and have worked independently for most of the time since. I co-founded Casting Frontier in 2006 and served as its technical lead through its 2020 acquisition by Talent Systems; the platform I architected included a database of more than 800,000 actors and supported commercial casting used, directly or indirectly, across much of the Fortune 100.
Alongside that work, I've written several books on software development, spoken at conferences internationally, and published technical articles distributed in more than 65 countries, in publications including Linux Magazine, Dr. Dobb's Journal, and ADMIN Magazine. Over the years I've worked with organizations ranging from Ivy League universities to small local businesses — a range that has taught me to meet a system, and a team, where they actually are.
Trusted by Leading Organizations
I've gotten a chance to work with a variety of industries, from Ivy League institutions to salvage motorcycle dealers and more.
Getting in touch
If you have a system that's slow, fragile, or hard to reason about, I'm glad to talk it through and give you an honest read on it. An initial conversation costs nothing and carries no obligation to go further. Being based in New England, I can often meet in person; video works just as well.
Call (603) 574-4766 Contact Me