Chris Harden’s Take on Operating Plans, Integration Risk, and Why Human Judgment Is the Last Thing AI Can Replace
Most engineering conversations focus on how to build better software. But the harder question, and the one that separates teams that ship from teams that stall, is how do you build the system around the software, meaning the people, the processes, the tools, and the communication that let a team actually execute?
Chris Harden has spent his career asking that second question. He started in sales before earning his electrical engineering degree at Auburn, then spent years programming control systems for venues including Epcot and Turner Stadium, and later worked on embedded software for products including the Coca-Cola Freestyle machine, Ford's infotainment system, and the Kindle Fire. From there, he moved into cross-functional game development at EA Sports, co-founded the Shark Tank startup TROBO, and later co-founded the Unity Orlando studio, where he now serves as Director of Production Verification for Product and Engineering and as an AI liaison helping bring AI-native products to market across one of the largest development platforms in gaming.
Chris recently joined EverOps' CEO, Stephen Koza, on Episode 7 of TechPod Talks to discuss what that range has taught him about leading engineering teams as AI reshapes the work. The throughline across all of it is a conviction that the durable advantages in engineering lie in the organizational and human.Â
Continue reading to get the framework Chris uses to lead that way, from the four pieces of any operating plan to the line he draws between what AI handles and what stays with people.
Every Operating Plan Needs Four Pieces (Most Only Have Three)
The most common way ambitious plans fall apart is that they look complete on paper and come up short in practice. The team gets confused, the tools fail to work together, and nobody can quite say what went wrong, because the plan was never wrong so much as incomplete. Chris codified the fix into a framework after watching the pattern repeat across reorgs, ground-up designs, and change management efforts throughout his career.
Most operating plans account for three pieces but miss a fourth. The four that actually matter are:
- People
- Tools
- Processes
- Communications
Holding all four consciously, and thinking about how they work together, closes the gaps that bite a team later when someone forgets a piece while moving fast. He built the framework first as a way to teach his own senior team at Unity the full shape of the production verification program he was standing up before realizing it deserved to be a book his team could hold in place of a scattered set of documents on a Google Drive. This later became his most recent title, known as The Proof.Â
The piece of the framework leaders most often underweight is communications. Chris is direct that it remains the single area where he coaches people most, from individual contributors to organizational leaders. The reason is that communication is a soft skill engineers are rarely taught and frequently assumed to lack. Proactive internal communication with the team, external communication with stakeholders, and external communication with customers all require deliberate attention, and the stakes become clearest during a reorg or reduction in force. This is where the combination of urgency and the need for early privacy leads to comms that rarely rise to what the moment demands.
Integration Risk Is the Hidden Cost of Coordinating People at Scale
Shipping a large product on a deadline turns coordination into the central engineering challenge, and the lesson scales from a two-person startup to a two-hundred-person game studio. Chris learned it most vividly leading development on an NBA Live title at EA Sports, a roughly two-hundred-person effort spanning multiple disciplines and locations, built to launch on the schedule that sports games live and die by.
The hardest part of an effort that size is getting that many people to converge on one cohesive experience. The mechanism that makes convergence real is the iterative build review. Every two weeks, the producer takes what each team calls done and puts it to the test. The work often integrates poorly the first time, and the definition of "done" has to be earned before anyone can truly claim it.
"It doesn't assume the worst in people. It just assumes that integration risk is always present. And that's something that often gets missed because no one's really in charge of integration."
The discipline Stephen described as trust but verify is the same one Chris runs on, and its value comes from treating integration as a standing risk with a named owner, which keeps it from drifting into an afterthought nobody holds.
That startup instinct actually predates EA, dating back to when Chris co-founded TROBO, for which he documented the full journey from zero to one through the pitch room and beyond in his startup memoir, Little Robot, Big Dreams. The book is now available on Amazon for anyone who wants an inside look at what that process actually feels like.
AI Accelerates the Work, But Judgment is Still the Human Edge
AI has moved engineering from a question of who knows how to use a chatbot to a question of who can build the back-end harnesses that spin up agents automatically, and the gap that once handed early adopters an easy advantage is closing quickly. Chris frames the shift through the lens of Geoffrey Moore's Crossing the Chasm, observing that the late adopters arriving now are stepping onto the other side of a leap the market has already largely made, with enterprise users building agent orchestration that his own team surfaces to him faster than he can keep up.
His relationship with the work reflects where the leverage now sits. Chris orchestrates agents more than he types code himself, defining the outcome, feeding it to agents he has built, and verifying the result over and over, which is the operating posture he describes as being the orchestrator and the verifier. When he used Claude Code to help write his most recent book, he built a set of specialized agents, including a structural agent that enforced his own rules, an editor, a quote researcher that flagged misattributed quotes before they reached the page, and a quality reviewer. The writing agent did less of the work than people expect because the redundancy and bloat that show up as AI slop required him to step in repeatedly and cut sections that restated the same point, until the result read cleanly.
The line he draws between AI and human work comes down to judgment and taste. AI is probabilistic, and when it makes a wrong call, it commits to that call at full speed with complete confidence, which is exactly where human judgment becomes a durable contribution. In game development, the version of this is taste, the designer's sense of the experience they are building, which AI cannot supply on its own.
"It's a pairing of the human and the AI that really works best."
That pairing extends to the broader debate over AI in gaming, where Chris places the creative and engineering sides on different risk profiles. On the creative side, AI-generated art, music, and models work well for rapid prototyping, allowing a team to check whether a game is fun before paid artists and designers produce the assets that go to market.Â
On the engineering side, agents deliver real efficiency for migrations, ports, and aging tech stacks, with the caveat that a leader has to be technically savvy enough to recognize whether the agent is producing slop that turns into years of tech debt, a risk that runs highest in mature products and lowest in greenfield work where teams move fast and shape the result as they go.
Communication Is the Skill AI Makes More Valuable, Not Less
Clear communication has always been the multiplier for technical work, and Chris traces his appreciation for it back to his earliest role in sales, where simplifying a message was the difference between a fast conversation and an onboarding that never landed. The skill he learned selling is the same one he now builds into how his team operates.
He turned it into a tool by building a jargon-removal agent that takes a PowerPoint deck, a document, or an email and returns a simpler version, which he has used to rescue roadmap presentations handed to him as piles of dense technical bullet points. The agent helps precisely because the underlying skill still has to exist in the person using it, the way a calculator helps only someone who already understands the math. But knowing what good communication looks like is the prerequisite for getting good communication out of an AI. The leaders who can connect with the person on the other side of the table, putting a message in terms that match what that person actually cares about, hold an edge that AI amplifies and cannot supply on its own.
"It's just like having a calculator. You have to know the math before you use the calculator. You have to know what you're doing before you get the right stuff out of AI."
The same logic shapes how Chris thinks about the next generation entering an AI-saturated profession, where the open question is whether defaulting to AI builds skills or quietly substitutes for building them. His own answer is to teach the fundamentals directly, which is why his kids run a summer business every year, scaled to their age, that puts them through door-to-door sales, packaging, branding, and the financials. In turn, this helps build the confidence and communication skills that no classroom curriculum delivers as well as real customers and hands-on experiences do.
How to Apply These Frameworks This Quarter
The conversation maps directly onto the decisions engineering leaders are making right now. A few concrete places to start:
- Audit your operating plan for the missing fourth piece: Take a current initiative and check it against people, tools, processes, and communications. The piece you have not explicitly planned is the one most likely to create a gap later, and communication is the most commonly missed.
- Assign an owner to integration risk: Identify who is accountable for ensuring the pieces fit together when multiple teams declare their work done, and institute a recurring build review in which done must be demonstrated before anyone claims it. Integration risk is always present, and it goes unmanaged when nobody owns it.
- Build a jargon remover for your team: Stand up an agent that takes dense technical decks, documents, and emails and returns a version targeted at a non-technical reader. The tool extends the communication skills that your strongest technical people may lack, and it pays off most in stakeholder and customer-facing moments.
- Match your AI risk tolerance to your context: Use AI freely for rapid prototyping and greenfield work where speed compounds and the cost of a wrong turn is low. Apply more scrutiny to mature products where agent-generated slop accumulates into years of tech debt, and keep a technically capable human to verify the output.
- Protect the judgment layer: Name the decisions in your workflow that depend on taste, judgment, and accountability, and keep them with the people who make them. AI does the work at speed, and the human contribution is connecting the ideas, making the calls, and standing behind them.
How EverOps Helps Engineering Teams Get There
The questions Chris explores in this conversation are ones we work through alongside engineering leaders every day. Helping teams build the operational systems around their software, aligning communication and process with the pace of AI-accelerated delivery, and drawing the right line between what agents handle and what stays with people are all central to how our team engages with its partners.
Our AI Opportunity Assessment is a strong starting point for teams mapping where AI can accelerate the work and where human judgment needs to stay in the loop. For organizations in the middle of a broader transformation, strategy consulting and embedded operations engagements provide the hands-on support that keeps engineering moving while the surrounding system catches up.Â
If your team is navigating any of the territory covered in this episode, reach out to our team today to start the conversation.
Keep Up With TechPod Talks
Chris Harden joined EverOps CEO Stephen Koza for Episode 7 of TechPod Talks. Subscribe to listen to the full conversation on Apple Podcasts, Spotify, YouTube, or the EverOps Podcast Page now. Follow Chris's writing at chrisharden.com and on LinkedIn so you don’t miss any of his insights. Be sure to follow EverOps for more information and exclusive updates from the series.
Frequently Asked Questions
What is TechPod Talks?
TechPod Talks is a podcast hosted by EverOps CEO Stephen Koza featuring candid conversations with technology leaders, engineers, and operators. Each episode explores how real teams build, scale, and operate modern systems, with a focus on practical takeaways.
What topics does Episode 7 cover?
Episode 7 features a candid conversation with Chris Harden on building the system around the software, the four pieces of any operating plan, why integration risk needs an owner, how AI shifts engineering toward orchestration and verification, where human judgment and taste remain the edge, and why clear communication becomes more valuable as AI spreads.
Who is Chris Harden?
Chris Harden is the Director of Production Verification for Product and Engineering at Unity, where he also works as an AI liaison bringing AI-native products to market. His career spans control systems engineering for theme park venues, embedded software on consumer products, game development at EA Sports, and co-founding the Shark Tank startup TROBO. He is also a published author whose books include The Proof, Mastering Commitment, and the startup memoir Little Robot, Big Dreams.
What is the four-piece operating plan framework?
The framework, which Chris details in his book The Proof, holds that most operating plans account for three of four essential pieces and that incomplete plans struggle in practice. The four pieces are people, tools, processes, and communications, and planning all four consciously, along with how they work together, prevents the gaps that surface later when a team moves fast.
Where can I listen to TechPod Talks?
TechPod Talks is available on Apple Podcasts, Spotify, YouTube, and the EverOps website, with episodes released in both audio and video formats.
Can I suggest topics or be a guest on the podcast?
Yes. You can share topic suggestions by reaching out on LinkedIn or through the EverOps website, which includes a guest request form for speakers interested in joining future episodes.
How does this episode connect to EverOps' work?
EverOps helps engineering leaders navigate the same questions Chris covers in this episode, including AI adoption strategy, organizational alignment, and the operational systems that let engineering teams turn AI acceleration into reliable delivery. Services like AI Opportunity Assessment, strategy consulting, and embedded operations map directly to the work Chris describes.
