CEOInterviews.AI
Start App
Ryan Petersen
Founder, Chairman & Chief Executive Officer, Flexport

Flexport 2026 Fall Technology Release Live Broadcast

📅 Sep 29, 2026 Flexport 36 MIN 16 SEGMENTS · 4 SPEAKERS
Ryan Petersen, founder and CEO of Flexport, unveils the 2026 Fall Technology Release: new AI and technology products built to ...

What Ryan Petersen said

Written from the verified transcript and checked against it. Every figure links to the moment it was said.

Ryan Petersen, CEO of Flexport, announced a suite of AI-driven products and updates at the 2026 Fall Technology Release. He emphasized AI's role in making supply chains more efficient and trustworthy, citing 150 operators in an AI accelerator and a run rate of 21 million automated tasks. New offerings include Flexport's MCP integrations with major AI tools, a rebuilt rates management system, AI-powered quote and booking agents, and a compliance audit tool that found a 20% error rate in competitor customs entries. He introduced the AI customs classification engine and the business rules engine for deterministic SOP compliance. Atlas was updated with multimodal routing, emissions tracking, and a disruptions layer. Flexport expanded to eight new countries, launched fulfillment centers in Canada and the UK, and opened a Chicago facility with 350 robots. Petersen reiterated the vision of an autonomous supply chain network.

Key takeaways

  1. Flexport is on pace to automate 21 million tasks at run rate, covering a significant portion of freight forwarding work.
  2. Compliance audit found a 20% average error rate in competitor customs entries, compared to Flexport's own 0.2% error rate on the same entries.
  3. Flexport's tariff simulator is the most widely used US tariff calculator, with CBP as its number one user.
  4. Over the past five years, Flexport has returned more than a billion dollars in tariff refunds to customers.
  5. Flexport opened a new automated fulfillment center outside Chicago with more than 350 robots.

Numbers and commitments

FigureWhat it refers toTypeAt
150 Operators in Flexport's AI accelerator metric 0:00
21 million Tasks automated at run rate per year metric 8:48
20% Average error rate in competitor customs entries audited metric 13:06
0.2% Flexport's error rate on the same customs entries metric 13:06
53 Changes to the tariff code in the last year metric 13:06
18,000 HTS codes in the United States metric 18:18
8 New countries Flexport has opened in this year metric 30:39
11 Number of languages Flexport is translated into metric 30:39
350 Robots operating in the new Chicago fulfillment center metric 30:39

Chapters

  1. 0:00Introduction and mission
  2. 3:52Flexport MCP with AI tools
  3. 6:00Rates management overhaul
  4. 8:48Trust and safety for AI agents
  5. 11:22Customer spotlight: Comfort
  6. 13:06Customs technology suite updates
  7. 18:18AI customs classification engine
  8. 21:26Customer spotlight: Life Fitness
  9. 23:06Atlas multimodal and disruptions
  10. 30:39Global expansion and fulfillment
Ryan Petersen 0:00 ↗
All right. So, thank you. So, we've been building Flexport for more than a decade now, and I can tell you we've never seen anything like the last six months in terms of our ability to build high quality software and bring it to you quickly. And the reason obviously is AI and it's our software engineers are now coding by just managing this swarm of coding agents. But it's not just the engineers anymore. We actually have more than 150 operators at Flexport going through an AI accelerator that we've built to teach them how to code, to teach them how to build software securely, to make dashboards to oversee their work, to automate their workflows, automate their team members' workflows. And increasingly, what they're doing is going on site with our customers and helping you automate your workflows and showing you how to get the most out of these technologies. And so the time to launch has collapsed from quarters today. So we're really just very very excited to show you the latest tech today that we're launching. But before I get into that, I want to remind ourselves why we do this. And it comes back, we've always been a very mission-oriented company. Excuse me, I left my clicker over here. We've always been a very mission-oriented company. Really starting going back to this idea that we want to make global trade so easy that there will be more of it. We do this because well, trade is a good thing. It leads to prosperity. Your business is growing, economic growth, all that comes with that. And if we're honest, the last couple years have been anything but easy. It's been kind of a nightmare of constant tariff changes and volatility. It's like impossible to keep track of. And not just the tariff wars and the trade wars, but real wars. Piracy, cyber attacks. I mean, you go down the list. It's been a bit of a mess. And all of that just makes us more passionate about our mission. It's more obvious that we need to go and solve problems and actually creates a lot of opportunity for us to create value for our customers. So it just gets us more and more certain that we're working on the right things here. And we've always been from the beginning to be a very customer obsessed organization. Like we don't buy this idea that there's this trade-off between technology and relationships or between scale and high-touch customer service. We want to be a white glove customer service organization even as we scale the business. And AI has finally once and for all made it very clear that that trade-off does not exist because AI lets you be really high-touch, really customized and tailored and help customers in real time with their problems. So it's just an amazing world and we're getting much closer to this vision of this autonomous supply chain network where AI is making most of the decisions. It's detecting problems before anybody's aware of them, calculating fixes, and then implementing them and actually just going to create these recurring self-improvement loops in a supply chain. And for the last decade, some of our competitors have made fun of us or ridiculed us and said, 'Hey, like tech doesn't matter. This is a relationships business.' But even they are changing their tune. It's obvious now that this is the inevitable future where if you think AI doesn't matter, you're going to lose. And so that's really exciting for us and I think a good segue to introduce you to one of our first and most exciting products that we're announcing today, which is Flexport's MCP.
This for those who are not familiar, this is Model Context Protocol. This is what lets your AI agent talk to our AI agents and we're launching today with partnerships with pretty much all the major AI tools whatever you like. We've got Claude, Copilot from Microsoft, ChatGPT, Muse, as well as Grok, Grokbot. So you can connect your favorite AI tool and work with Flexport inside your tool. And so let me show you how that works. This is the Claude example, but it's very similar for the others. Go to the connectors menu inside your AI tools there and just search for Flexport and add us. And from there, you'll be able to do a whole bunch of powerful things. Starting in this example, just use natural language and you can track shipments, say, 'Where's my stuff? What's happening?' And we'll get you real time status updates for what's going on with your freight. Now you can also just talk to it natural language again and get quotes to ship anything anywhere by any mode of transport and these are bookable. So you can just actually book freight inside the AI tools. I think this is a really big deal. You know, that's a nice little simple user interface kind of friendly and nice but you're going to see really powerful automations that you're going to be able to build. Like I'm really excited to see how you put this to work because now of course your AI has this capability to go ship containers anywhere in the world to go execute and you can build all sorts of automations about this. The more data that you connect into the same AI tools you already use where you've got your enterprise data governance policies in place. So you can now if you add your inventory data which if you use Flexport fulfillment is also very easy to do. But you add your inventory data, you can connect email tools to be able to set up automated workflows. So in my view, this is not just a really powerful kind of moment in logistics industry, but for the AI industry, like allowing AI to ship containers around the world is kind of a big deal. It's going to be very exciting to see where this goes. And so that's quoting and booking inside your AI tools.
Well, we also made some big changes to how we do quoting and booking inside the Flexport platform. Frankly long overdue changes. We've totally rebuilt our rates management system from the ground up. Unifying air freight, ocean freight by all contract types. So spot market, whether you have a named account kind of contract for an annual basis or you're getting monthly rate sheets from us. These are all now in one friendly kind of e-commerce experience. Make it very easy to quote, book, check out like you'd expect from an e-com experience. It's a much better flow. Thank you for your day-to-day operators doing this work, for partnering with us, and it's very collaborative, so it works with your suppliers as well. So, that's an update to the rate sheets there. Many of our customers actually don't really want to log, they don't want to use the AI tool or they don't know it's they haven't gotten there yet. They don't want to log into the platform. They just want to get their rates in a nice Excel file. That's very common, very normal because you have your own templates. You want to see like the line items broken out a certain way and always have it be the same way. Historically, that's been a lot of work for our teams. We have to go and do that ourselves. So, with this update, now we have an AI agent that will transform our rates into your template automatically, very quickly, freeing up our teams to do more. It's not the highest calling of civilization transforming Excel files. So, this frees our team up to do more important strategic work for you. So, I think you're going to like this. It'll make us much faster and just better. Next, a lot of you still prefer the really old-fashioned way. Just email and get some rates. Sometimes that's easier, frankly, right? Just email us. So, we've now built an AI agent that's monitoring our team inboxes where you communicate with us and looking for quotes and booking requests and it will just automatically if it's a quote generate the quote for you and return that and let you know, put it into the platform and then send you an email letting you know it's ready. Sometimes instantly but always very quickly showing you the rates right there. And again, you can go right through that same e-commerce experience to confirm the booking or you can just send us a booking and we'll by email we'll take care of it. I think all this makes clear that logistics now has a new workforce and I'm only just showing you a couple of examples above the surface here. We've built hundreds of these types of AI agents that are automating workflow at Flexport, freeing up our teams to do more strategic work. And this is an infinitely scalable workforce. They don't sleep. They don't take vacations. They improve quality and efficiency.
And but the interesting question I guess you know as we've done this because we've automated we're on pace now run rate to automate 21 million tasks. It's a huge percentage of the work of freight forwarding and logistics is being taken over by AI agents. And the question we ask constantly and the thing that we focus on more than anything else is how do you trust these agents to do a good job? Because we all know these are probabilistic systems. They're prone to hallucination. It's getting better as these models continue to improve, but it's still a really big deal. And supply chains are too important just to yolo it and hope that the AI agents get it right. And so what we've built is there's a number of strategies that we use here to get to trust and safety as we do this. And that starts with eval. A set of tasks that we put through every agent long before it's able to go into production and actually do work in the real world on behalf of our customers is we're doing a huge gamut of tests and testing a thousand different ways to make sure that it gets the result that we want. We have a tracing methodology for reviewing the chain of thought for every agent for monitoring anything that's not working as we'd expected. We do a lot with humans in the loop where there are people monitoring this and especially early days when an agent's first coming online, we don't let the agent do any work. We have it working in parallel with the human operator doing that job. In some cases, we'll never take the human out of the loop, especially in compliance areas or in customs. And I'll show you some examples of that a little bit later. But ultimately, what this is doing is allowing our team to focus on more important things. We don't believe that this is about cutting headcount. I think you're going to see at Flexport's headcount grow faster as a result of being better at our jobs and having a higher quality service that we can grow faster. It's an interesting dichotomy because you talk to a lot of companies, they're just like, 'Oh, I can use AI to cut costs and remove labor.' But we think that's not the important thing that it can do. And the most important work we can do is for sure is help you grow your business. That's the main way that we can align our interest with our customers' interest. If you grow faster, you'll ship more stuff. We'll make more money. It's a happy world that we want to live in. Well, I want to hear I'd love to hear from one of our fastest growing customers is Comfort on how we've helped them with our technology and our solutioning.
Comfort Representative 11:22 ↗
Comfort is one of the fastest growing apparel brands of all time. We are wildly known to be successful for our fleece hoodies and our airplane mode hoodies specifically. We're pushing into other apparel categories as well as home. So, we now have the number one selling blanket as well.
Flexport Representative 11:40 ↗
I didn't quite realize what a phenomenon Comfort was until I told my niece that we work with Comfort. She couldn't believe that her uncle gets to work with Comfort. She couldn't believe that I could hook her up with a Comfort hoodie. So, that's when I realized, wow, there's more to this brand and the phenomenon that it's becoming. We have been so lucky to grow with Comfort over the years. Back when we started working together in 2022, we saw this incredible product that had really, really exciting things ahead.
Comfort Representative 12:10 ↗
Last minute surprises are never something that I want to experience, but when and in the off chance that we do, the Flexport team has been absolutely wonderful and very quick to get things done, to respond, and to move fast. And that's the flexibility that allows us to have those last-minute viral events. We've gotten really good at predicting with as much technology and predictive patterns as we possibly can. And then ultimately the sharing of labor within the warehouse allows us to handle the spikes that can feel unpredictable. I love working with Flexport. I love a partner that's able to bring me an issue before I find it myself. There's no way we would be able to do this without the relationships that we built with the team across the entire Flexport organization.

5 more exchanges in this transcript

Sign in free to read the rest of this interview. No card required.

Sign in to read the full transcript

Cite this transcript

APA, MLA, BibTeX
APA

Petersen, R. (2026, September 29). Flexport 2026 Fall Technology Release Live Broadcast [Interview transcript]. Flexport. CEOInterviews.AI. https://ceointerviews.ai/interview/2949256/

MLA

Ryan Petersen. "Flexport 2026 Fall Technology Release Live Broadcast." Flexport, 29 Sep. 2026. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/2949256/.

BibTeX
@misc{petersen2026_2949256,
  author       = {Ryan Petersen},
  title        = {Flexport 2026 Fall Technology Release Live Broadcast},
  howpublished = {Interview transcript, Flexport. CEOInterviews.AI},
  year         = {2026},
  month        = {sep},
  url          = {https://ceointerviews.ai/interview/2949256/},
  note         = {Speaker-attributed transcript with timestamps}
}