Jamie Schiesel

Excited to share that I've been promoted to Senior Director of Project Management at Teal Media. Grateful to work somewhere that keeps pushing me to grow and sharpen my craft. And even more grateful that they know exactly what I need on my promotion photo: pitbulls. Lots of pitbulls. Here's to more great work, great people, and definitely more dogs ahead. #TealMedia #ProjectManagement

“An agent is just prompts” is the dumbest sentence I have heard in 2026. It sounds clever at a conference. It collapses the second you try to run a real business with one. A prompt is a sentence. An agent is a system. What actually makes an agent work: 1. Context that turns a generic answer into the right one 2. Tools that act in the world, not just describe it 3. Skills that make execution reliable 4. Sub-agents that handle bounded specialist work 5. Verification that catches “confidently wrong” before it ships 6. Compounding memory that makes the system better with use The constraint was never model intelligence. It was whether anyone built the supporting structure to activate a role based agent. If your “agent strategy” is still mostly prompt craft, you are optimizing the doorbell and ignoring the house. I wrote the full end-to-end definition of what an agent actually is. Link in the first comment. My PoV. Share yours in the comments. [sidebar - the Linkedin writing agent tried to soften my leading sentence, I rejected it because it should not be softened]

Most companies are not short on AI ideas. They are short on operating clarity. This article breaks down why AI sprawl happens when teams spread across tools, pilots, and experiments before choosing the first workflow that should actually change. Read the full blog here: https://hubs.ly/Q04nwvRR0

Last week, I had the pleasure of presenting at the C-Vision International data and marketing think tank with Taylor Fisher . It was a fantastic opportunity to connect with a remarkable group of leaders and discuss the changes in marketing over the past 18 months. I would like to extend my gratitude to the C-Vision team and everyone who participated. I truly appreciated the chance to be part of this insightful event.

Yesterday I watched my son fail. A dozen times. On the same move. Same spot on the wall. Over 10 minutes. Fall after fall until he committed to the dyno. Threw for the hold and stuck it. I love rock climbing. As a sport, it's brutal and honest. A gym owner opened our 2026 Divisionals training with a line I keep coming back to: 👉 "Climbing fails 99 out of 100 times." Walk into any gym. You won't see people floating up every route. The harder the climb, the more they fall. And looking around, you see falling everywhere. My son climbs hard. 12Cs. 12Ds. He's chasing a 13A this summer. (that's really hard) He falls more than he sends. Here's what I've learned watching him: Every fall is information. - A different foot placement - A new hand sequence - More power into the next move - The nerve to go for the last hold, knowing he might drop all the way down He doesn't fall and quit. He falls, he learns, and adjusts. Most of us don't. Somewhere along the way, we learned to avoid the hard parts. - We say we can't - We give up - We choose to do something easier That little voice telling you you can't? We all have it. I do too 🙋‍♀️ I've heard it through jobs that didn't work. A business move that flopped. Relationships. And I hear it now, building my own agency from scratch. The voice isn't the problem. What you say back to it is. So next time you're stuck on the same move for the tenth time: - Name it. "This is hard", not "I can't." - Change one thing. Not everything. One. (a foot, a grip, your mindset) - Separate the fall from the failure. Falling is data. Quitting is the failure. - Talk back to the voice. Out loud if you have to. - Remember the last hold. The biggest move usually comes right before the send. My kids taught me this without trying. There's so much joy when they finally stick it. Pride. Relief. And then? They look up at the next route. Maybe that's the whole thing. You send, you celebrate, you find the next wall. What move have you been stuck on?

When teams evaluate new technology, most of the attention goes to what's easy to see: the demo, the integrations, the feature list, the roadmap. The questions that matter just as much often get far less attention. Who has access to my data? How is it stored? What happens when something goes wrong? Security and compliance are often treated like boxes to check during procurement, but they shouldn't be. Protecting customer data is not a differentiator, it's the baseline. Compliance and security are not the same thing. Compliance means you've implemented the right controls and can demonstrate that they're working. Security is whether you're actually protected in the real world. You can pass an audit and still have vulnerabilities. Compliance is a point-in-time validation and provides structure. Security is an ongoing discipline, and requires continuous attention, judgment, and improvement. I learned this firsthand early in my career. I worked with a team that operated under a pretty common assumption: we were too small, too niche, too unglamorous to be a target. We found out the hard way that ransomware doesn't care about your industry or your headcount. Rebuilding from that experience was painful, but it also permanently changed how I think about security. "Who would bother with us?" is never a strategy. That mindset shapes how we operate at MetaField. We build software that sits at the center of geotechnical and construction materials testing workflows (boring logs, field inspections, lab results) and we take that responsibility seriously. Security isn't something we layer on after the fact. It's part of the foundation of everything we build. We view compliance frameworks as a starting point, not the finish line. Achieving SOC 2 is important, but the certification itself is not the goal. The real value comes from the process behind it. It forces us to challenge assumptions, strengthen controls, refine processes, and think more intentionally about risk. That work doesn't stop when the audit ends. It shows up in the conversations we have during design reviews, not just audits…and in expecting every engineer to think like a defender, not just a builder. Most importantly, it means asking the hard questions before our customers have to. The trust our customers place in MetaField isn't built on a certification badge. It's built on the work that happens daily between audits. If you're evaluating a software platform (in any industry, but especially one where your data underpins real-world decisions) here's what I'd push vendors on: what their security practice looks like between audits. Ask about their threat modeling process. Ask what happens when something goes wrong. The answers will tell you a lot. I'm curious how other technology leaders approach this. How do you keep security from becoming an annual checkbox?