The AI Tool Helping Warehouse Teams See the Messy Middle

In this episode of Everything is Logistics, Blythe talks with Sarit Tamir, CEO and co-founder of Seeteria, about how computer vision is helping warehouse teams find lost time inside their operations.

SeeTeria is a software-only company that connects to existing CCTV cameras inside warehouses, distribution centers, fulfillment centers, and 3PL facilities. The platform watches floor activity, detects operational friction, and sends real-time alerts to teams before small problems turn into bigger delays.

They cover:

• Why warehouse teams still miss what happens between system scans
• How existing CCTV cameras can become an operational visibility tool
• What dock queues, idle doors, dwell time, staging congestion, and near-misses reveal
• Why the “messy middle” of warehouse operations adds up quickly
• How real-time alerts help floor teams act faster
• How managers can use shift summaries to find recurring patterns
• Why the goal is to support warehouse workers, not replace them

This conversation is part of the CargoRex AI Use Cases in Logistics guide, featuring real examples of how logistics companies are using AI across freight, warehousing, procurement, visibility, and operations.

Read the full guide here:
https://cargorex.io/research/ai-use-cases-in-logistics/

LINKS:

SeeTeria:
https://seeteria.com

CargoRex AI Use Cases in Logistics Guide:
https://cargorex.io/research/ai-use-cases-in-logistics/

-----------------------------------------

THANK YOU TO OUR SPONSORS!

SPI Logistics has been a Day 1 supporter of this podcast which is why we're proud to promote them in every episode. During that time, we've gotten to know the team and their agents to confidently say they are the best home for freight agents in North America for 40 years and counting. Listen to past episodes to hear why.

CargoRex is the search engine for the logistics industry—connecting LSPs with the right tools, services, events, and creators to explore, discover, and evolve.

Digital Dispatch maximizes and manages your #1 sales tool with a website that establishes trust and builds rock-solid relationships with your leads and customers. 

Transcript

SPEAKER_00 0:05

When we walk with uh let's say a warehouse manager that knows that he's currently paying too much detention fees and he's not keeping out uh keeping up with their schedule. Walk with this person, understand, okay, that's the threshold, that's one you want to measure. Let's do that. Let's walk for one week, have a baseline, and then start providing you the information. To your team, team, we walk with the team to make sure that they understand how we are there to help them. We are not there to replace them.

SPEAKER_02 0:42

Welcome into another edition of our Cargo Wreck series where we are talking about AI use cases in logistics. And today we are talking with Sarit Tamir. She is the CEO and co-founder of Ceteria, where they are attempting to eliminate the messy middle in warehouse operations. And not just attempting, but actually doing that. We're going to talk about those customer stories a little bit later on in the episode. But first off, for I guess first and second, I want to welcome you to the show. And then also I want to hear more about where the company named Ceteria came from.

SPEAKER_00 1:14

Yeah, of course. And thank you so much for having me on the podcast. And that's really exciting. Uh about our name. Actually, it's quite a family experience started at. We were at uh in a trip to Rome, and of course, we visited the Vatican, and there was one statue of a goddess that really caught my eye, and I went over and checked, and it said that this is Soteria, and she is the goddess of preservation and resilience.

SPEAKER_02 1:47

That I love that name, and I'm I'm trying to be objectional and or uh in these types of interviews that we're doing for cargo wrecks, but you told me that story before we started recording, and I was like, I have to hear it. It it gives me goosebumps to hear it because I didn't even know that that goddess existed. But what a great name, what a a great sort of uh, I guess, role to or that symbolism to take into this industry.

SPEAKER_00 2:11

That's what what we are trying to do. We just made a small adjustment and changed the so in the beginning into C to reflect what we are doing, which is using computer vision to achieve achieve those goals. That's the start that's the story, my story behind the name Ceteria.

SPEAKER_02 2:31

Well, I love thoughtful business names because it it shows that you thought really long and hard about it. And then also that could you know kind of cue into what you do within your business is seeing more and and fixing that that you know the what do they call it, the messy middle of of warehouse operations. And that's what I was reading from your website. And it's it was things that I didn't even really consider from you know monitoring, you know, uh, you know, workloads and I guess let me have you tell the story of what your company does at a high level and in your target audience or target customers.

SPEAKER_00 3:08

What Citarian is doing is giving the warehouse teams real-time visibility into how work is actually flowing on the floor. We use the existing CCTV cameras you already have on site. Sometimes people are confused and they're asking me if we are selling security cameras. We are not. We are a software-only company, we use the cameras that you already have on site. With them, we observe the activity on the floor, and we turned those movements into operational signals, like queues at the docks, idle open doors, staging areas that are filling up, dwell, loading, loading poses, and near-miss interactions. They all seem like small events, but everything when everything adds up, it's a lot of time that is being wasted. So, what do we do with this information? The floor on the team receives those signals as live alerts in real time to uh a mobile app. So they know when something is happening and they need to act really quickly to fix that before it becomes a major, a major issue on the floor. And the managers, the managers can go into the system after the shift, investigate, know exactly what has happened, get all the information that they need. Uh because what we find is that this information is uh is currently missing. Um you know, everybody is using WMS. We are not here to replace that, but uh this uh gap, this invisible middle that you mentioned is what's happening on the floor between the scans.

SPEAKER_02 5:03

And so for a lot of that, I I guess that I guess paint the picture of what uh a bad operation looks like. What were what are some of those uh alert examples that a manager would be getting that could cause future issues?

SPEAKER_00 5:19

Okay, I I don't think that um I don't think that these are signal of uh facilities that is not uh well uh operated. Because even if the with uh the facility the facilities that we worked with, even those that had uh very experienced teams and were using software, still a lot of events that were unexpected, unplanned were happening. And for example, when uh a truck comes over, the it the uh it connects to the door, the door is open, but nothing happens. Falklifts are not coming, no walk is being done, and sometimes they come over, they drop some pallets, and they just go. And it's sometimes really easy to say to the supervisors, you had to be there, but most facilities are very big. And the people on the floor, you know the statistics, uh they say that they they walk or run about five to ten miles each day, so nobody can be everywhere all the time. So these uh things uh go unnoticed even in in the best managed facilities.

SPEAKER_02 6:53

So it almost sounds like it these uh little minor things that you wouldn't think too much of, they compound over days and weeks and even months.

SPEAKER_00 7:04

Yeah, and and even in in one shift, because you know sometimes it looks like, oh, okay, it was just five minutes. Or maybe it uh you know, the staging area was filling up. And yeah, we know that the fault lifts were shuffling their way to find the palette they needed, but it didn't take long. But when we measure that in the system, and as you said, we add everything up, it's hours.

SPEAKER_02 7:33

And then those hours are obviously affecting you know bottom lines and profit and loss statements, and that compounds over time. And so for your business and for your software, you plug directly into the the already existing cameras. Is there you know a set amount of cameras that typically exist in a warehouse? Or is it you know are there certain cameras that provide more value versus other cameras?

SPEAKER_00 8:02

It never happened that we uh came to a facility and the cameras were not good enough.

SPEAKER_02 8:08

And so when they're not good enough, then that's when you have to make the suggestion.

SPEAKER_00 8:12

Sorry, I didn't I didn't explain it well. All standard industry cameras uh are completely uh fine to be used with our software.

SPEAKER_02 8:22

Okay. And so once they walk me through what I guess a a typical onboarding process looks like. What kind of customers I guess are a good fit, and then how do you make sure that they're taking full advantage of the software during their onboarding process?

SPEAKER_00 8:39

So uh customers we are aiming at are warehouses, uh distribution and fulfillment centers, uh 3PLs, usually medium to large. Well, you mentioned before that sometimes people don't even know that the problem exists. So for us as a startup, it's really important at that point to work with companies who understand there is an issue of lost time. And then the onboarding from our side is is quite simple. The the hard work from our side was already done when we developed the software, and it was really important for us uh to create a solution that is really simple to bring in, to implement, and and to use, because it's all about the customer experience and the value. So we come over what we need is a Wi-Fi connection to the cameras, okay, and internet connection to send the data, the process data to the cloud. It usually uh takes two to three days to set everything up, even less. And uh yeah, and then because we do not connect uh to any system, there's no integration.

SPEAKER_02 10:04

So it makes it the sign-up probably very very uh simple and maybe within days that you can get set up.

SPEAKER_00 10:12

Yeah. Uh usually it takes a day or two. So we come over, uh we uh connect to the cameras, and then we start um processing to create a baseline. Because at that point, the most important thing is to show value. So to see an improvement, first we uh create a baseline baseline. So we uh measure the KPIs of all the friction points where time is lost, where walk is stuck, okay. Uh where I I also mentioned the near misses, and what I mean is uh a near miss of a forklift, uh forklift and the person. Okay, so we measure all these friction points, and then we bring on board the users and we keep measuring. And then what you see is once they start using that, and a supervisor, although he's somewhere between the aisles, uh making sure that his picking is going correctly, he currently doesn't know that there's a queue of Hawkly foaming at one of the doors. But the cameras do see that. So at that point, he gets an alert on his phone telling him he needs to go there. So just think about it. At that moment, we uh saved time uh that improved the the the doc utilization and prevented potential fines because of you know um the driver's detention. And when you look at uh all these alerts, real-time alerts, and the ability of the supervisors of the team to be in better management of their time, you get also improved labor management, so much walk done, done in in minutes that otherwise would have been lost.

SPEAKER_02 12:30

Do you find that a lot of your your customers don't even know it's a problem yet? Or maybe they have some of these issues until they start using your software where it's plugged into the entire warehouse. It it almost feels like they, you know, they don't know what they don't know, but when they do know, they can address that problem and make these incremental improvements.

SPEAKER_00 12:54

Up to it, this is a real marketing issue. Um this is not something that a manager who is looking to improve will go into Google and ask, is there a vision system that can help me save time? But when I I speak to people and I describe the issue, everybody's saying, Yeah, yeah, we know we have that. But we are trying to solve that with getting more experienced people and hiring more people. And some of the the issues they say, okay, so we didn't just didn't think that we can improve that. So that's that's a gap. That's a gap of of time that we can now provide with computer vision, with image processing. And I'm gonna say it only once because I hate that people overuse it with AI. So we we bring this lost time back.

SPEAKER_02 13:56

Well, I think that that's a really good use case for AI. And I I know for a lot a lot of folks they they have some exhaustion around the phrase. I think it's probably used in in situations where it doesn't make a whole heck of a lot of sense. Um, but people have used AI for years. I mean, if you use Google, if you use Spellcheck, you know, things like that, that's all AI. And so I think what the AI is, is what's happening here, and what you're describing is that we can be smarter about things that we didn't know was a problem and didn't know existed, where even the most experienced managers uh might have these gaps because they're they can't be everywhere at once, but this software allows you to be everywhere at once within the warehouse, if I'm understanding correctly.

SPEAKER_00 14:42

Yeah, exactly. Exactly. Uh there's uh yeah, actually there's an example that I really like. Uh one uh one person told me, I feel like I'm currently have the superpower of teleporting.

SPEAKER_02 14:59

Yeah, that I I mean I could definitely see where, especially from a manager perspective, of where they would need to go. You have probably a set amount of things that you do every day, a routine, and then you you don't know it's a problem until the end of the shift. Whereas it sounds like with your with your system, now and correct me if I'm wrong, is this a separate app or is it an app that sits inside of the WMS? Or how how do you interact with the alerts?

SPEAKER_00 15:28

No, we walk along the WMS. We do not integrate with the WMS.

SPEAKER_02 15:34

Okay, and so how what does that look like? Is it a you know a a a bookmark on your phone or uh you know a browser tab? It's an app. Oh, okay. So it's an app that you get push notifications whenever something is going on.

SPEAKER_00 15:48

Okay, so there are two types of users uh uh to to the Citeria app. Uh what I mentioned until now was the the live view that is usually used by the floor team. So they get the notifications in in real time, and they can also communicate with the other team members. So it won't happen that two people are gonna go and uh uh uh attend with the same uh the same event. So this is for the the floor the floor team. But now let's let's think what's going on with the managers because when the shift was completed, the manager can open the app and okay, they can also run it on the on the computers on the browser. And then they can look at the summaries of everything. And for example, to see where there were recurring events. Okay, so let's say that you are a manager and you uh open the uh the shift summary and you want to see what happened last week on Wednesday on the morning shift. So you get a summary of the events. So if, for example, you see that in staging area number four, the it was filling up, um it was filling up at uh 10 in the morning and then one hour later. And so you see a pattern, you see that something is happening that needs your attention. Okay, so you you get all the information of what happened, all the execution, um, all the execution data that now you can use to better improve how things are being managed on the floor. It's really important to give the people on the floor the information in real time, but uh the people who are responsible for improving the processes and identifying problems are the managers.

SPEAKER_02 17:49

Gotcha. So so two different use cases with within the same warehouse, but still providing that incremental value. Where does so it it sounds like uh uh you know you you have your system and then how does AI play a role on top of it? Is it you know giving recommendations in the summary? Is it you know maybe a a chat box where you can have a conversation with it? Where does AI play a role?

SPEAKER_00 18:16

Well, uh since we are not doing uh LLM and that's not um uh any generative AI, we work differently. What our uh AI is doing is is doing the analysis of the video stream. Our input is not data and is not uh is is not letters, our input is the video. So when we analyze that, we identify the object and we identify the behavior and we predict uh where what kind of activities are about to happen. So when you have uh the trajectory of the objects and you have the the location of them and you have your history data, then you are able, even if you see just uh let's say um two forklifts at at a dock, but you see other forklifts moving at that direction, and that's something that you saw happening in the previous days. The system is able to identify that an issue is about to happen and alert the manager when the problem is still small. So, yeah, that's that's the AI part.

SPEAKER_02 19:37

Gotcha. And so it sounds like it it could be a system of not only performance monitoring, but able to catch something before it becomes a larger issue.

SPEAKER_00 19:48

Yeah, exactly. That that's the that's the training of the algorithm and that's the predicting part. And in addition, uh what I just what I described um before was the the shift view where the manager can go back and see everything that has happened and see the connections between the events and uh all how all the KPIs are presented over there, but that's not all. Because once you have the information in the system, and yeah, that that will take um several months when you have everything now the system uh starts um um um recognizing patterns and anomalies and making all the connections so uh you don't have to make them the yourself. The system actually surfaces everything and provides you with all these insights so you can change things in the process, in the training, in understanding where you have uh too many uh forklift or people walking in one part of the facility and not enough in the other.

SPEAKER_02 21:05

And so then you're you're you're gathering these insights and then the or the company is gathering the insights and then can make those actionable decisions. And and you know, with uh a lot of the KPIs that were mentioned, I think idle time, Q minutes, congestion duration, dwell time, recurring flow patterns. I would imagine that this surfaces uh a lot of trends for these companies to be able to take action in the future. And I believe you had a you know a pretty interesting use case um with one of the larger customers probably in the world. Do you want to tell us a little bit about that?

SPEAKER_00 21:41

We worked with Coca-Cola last year. We uh validated our technology with them, but the use cases were different. Okay, they were customized to their internal needs. So it's uh it they're not the use cases that we are speaking about now. But what it gave us as a as a um software house, as a as a software vendor, is the verification that the that the algorithms are are excellent and are providing exactly what they should do.

SPEAKER_02 22:17

So it almost sounds like that was uh the the Coca-Cola case study was almost a baseline for your software. And then maybe something that you know some other players or some other customers could learn from for you know a company that's been around for gosh, oh I think more than a hundred years. And so maybe the way that they do things is something that can is a good study of how you know other places can make improvements in their messy middle of warehouse operations. Yeah. So what um for for your is there a particular use case or a particular customer that is a best fit right now? Is it you know the the enterprise customers with you know hundreds of thousands of square feet of warehouse space? Um is it maybe some smaller warehouse operators? Uh what is a good use case for the Ceteria uh technology?

SPEAKER_00 23:09

I think that's uh it's it's less about the the size of the facility. Of course, yeah, it won't make sense to do any implementation for a facility that has like uh less than 10 doors, okay? But uh when we go uh beyond that, uh for us, it's more important who are the people that we are working with.

SPEAKER_02 23:33

Interesting.

SPEAKER_00 23:34

So sometimes in corporate the challenge is that the people who are um in charge of the implementation on site are not the decision makers. And when you want to do a really, really uh short pilot, you want to uh be able to walk with the people who can uh communicate with you without needing to go to get any approval, and also to be those that are concerned about the problem and are committed to solving it. Because when we walk with uh, let's say a warehouse manager that knows that he's currently paying too much detention fees and he's not keeping out uh keeping out with their schedule. So we can work with this person, understand, okay, that's the thresholds, that's what you want to measure, let's do that, let's walk for one week, have a baseline, and then start providing you the information through the shift to your team team. We work with the team to make sure that they understand how to use the system, although it's really simple. And he is or she is involved to to actually make sure that they are on board, understand that we are there to help them, we are not there to replace them.

SPEAKER_02 24:57

Yeah, I I think that that's a a concern for a lot of workers, or a growing concern really for a lot of workers is that if they even hear the phrase AI or technology, it it feels almost like either it's gonna take my job or it's a burden to learn this new system. But when you can show folks exactly what's happening, the problems that it's solving, and the problems that it's fixing, you know, maybe these things uh the the customer or even the warehouse worker just weren't even aware of. So it these incremental betterments only only get better if you actually know that they exist and the manager and community.

SPEAKER_00 25:41

Because if you know, if a worker will get an alert and they will ignore that, I can have the best technology, no improvement will happen. That's why it's it's it's crucial to to pilot in facilities where everybody is on board and really want to succeed, um, because there's there's a benefit to everybody. For the managers, of course, they want to save time, they want to save money, but for the the the workers as well, because you don't want to be stressed out the entire time to be running all over and always to to be um afraid that something will you know will will go unnoticed and you might lose your job. So we are we are there to to support them, to empower them. Um not to to replace anybody.

SPEAKER_02 26:46

And so you had mentioned earlier that it it only only takes a couple of days and a a Wi-Fi connection in order to uh you know sync in with their CCT or CCTV cameras. And you you mentioned the the establishing a baseline, you know, from after your your cameras are you know integrated or you know set up properly, what does the how long is the baseline process and then when do you expect to start getting some actionable feedback or actionable alerts?

SPEAKER_00 27:15

Well, it depends on the level of the activity in the facility, okay? Because we need data. So in average, it's about uh two weeks.

SPEAKER_02 27:25

Oh, okay. I I don't know why I thought it would be much longer, like months.

SPEAKER_00 27:29

Well, it depends. If it's winter time and there's uh you know uh very low activity, it will take more because we don't have enough enough data to train the algorithms and to create to establish a baseline. Yeah, if it's if it's summertime and there's a lot of activity. And especially actually we love summertime because when not all the team is is uh the experienced uh guys and there are a lot of uh seasonal workers and everything is so much more hectic and difficult to manage, that that's where we we show value uh very quickly.

SPEAKER_02 28:07

Yeah, that that makes sense. And so then now that you know with the full understanding of you know the implementation process, which sounds very you know, low barrier of of entry, and then fixing those those things that could happen, not just on a one-time basis, but ongoing, that could be addressed maybe in in future employee onboarding and future employee trainings. And so those savings just compound over time. And I imagine you're the the safety scores within your business also increase. And then maybe that could lead to, I know at least on the trucking side of things, if you're a a safer carrier, then that leads to cheaper insurance. And so I would imagine that maybe that that has some ripple effects within the warehouse, too, is that if you have uphold a certain safety standard, then your employees are happier, they're getting the job but done more efficiently, and then uh maybe you can save some money on insurance costs as well.

SPEAKER_00 29:03

Yeah, for sure.

SPEAKER_02 29:05

All right. Well, I think that that's a a good place to leave it. Well, a couple more questions. Um, anything that you feel is important to mention that we haven't already talked about?

SPEAKER_00 29:15

Not at the moment, but I'm sure that once we hang up, I'll think of something.

SPEAKER_02 29:20

If only we had, you know, the the the soteria alerts for ourselves. So we can, you know, have that. Maybe that that's a you know a future product idea um for for you guys is that you know we can have our own personal soteria that that tells us and you know when we can be more resilient. That's one thing I need to consider. It would be a lot, you'd have to have access to a lot of cameras and data. So that's the only caveat. But um just just joking. Um, but uh all right, so uh Suri, where can I send folks um get signed up for a demo, talk with you? Um, I'll add all of those links in the show notes.

SPEAKER_00 29:56

Uh on our website, there's a link to the contact page. Just leave your um you know details and we'll be we'll be in touch with you.

SPEAKER_02 30:06

Okay, perfect. Well, well, I will make sure to put that in the show notes just to make it easy for folks. But Zarit, this was uh an awesome conversation. Thank you so much for joining us.

SPEAKER_00 30:15

Yeah, thanks so much for having me.

SPEAKER_01 30:21

Thanks for tuning in to another episode of Everything Is Logistics where we talk all things supply chain for the thinkers in freight. If you like this episode, there's plenty more where that came from. Be sure to follow or subscribe on your favorite podcast app so you never miss a conversation. The show is also available in video format over on YouTube just by searching Everything is Logistics. And if you're working in freight logistics or supply chain marketing, check out my company Digital Dispatch. We help you build smarter websites and marketing systems that actually drive results, not just vanity metrics. Additionally, if you're trying to find the right freight tech tools or partners without getting buried in buzzwords, head on over to cargo rex.io where we're building the largest database of logistics services and solutions. All the links you need are in the show notes. I'll catch you in the next episode and go Jags.