The BUZZ @ SWARM - Episode 1

About This Episode

We’re excited to publish our inaugural SWARM Engineering Podcast! Our first guest is Claudia Roessler, Director at Microsoft Cloud for Industry. You can listen along as they talk about about sustainability and solving problems in the food value chain, using AI data science and machine learning.

Transcript

Holly: Thanks for joining us today for episode one. Yes, that's right. This is our inaugural show on the SWARM engineering podcast, where we talk about solving problems in the food value chain, using AI data science and machine learning. I'm Holly self VP of marketing at SWARM engineering. And I'll be your co-host today.

 If you're curious about using data science and analytics to predict issues in the food value chain or how companies are using data to tackle sustainability initiatives. Then this podcast is for you today. We have a fantastic guest on our podcast, Claudia Roessler from Microsoft. Claudia is director in Microsoft cloud for industry and responsible for strategic partnerships in food and agriculture. She helps organizations on their digital transformation journey and supports them in building digitally enabled and data science solutions in the agriculture and food value chain. She's an ambassador for women and food and agriculture and serves as venture partner for radical growth. 

Today, Claudia is joined with your host and SWARM’s CEO and founder, Anthony Howcroft. Welcome to the show.

Anthony: Thanks, Holly. And hi, Claudia. Great to have you here on our first SWARM podcast. How are you? 

Claudia: I'm doing well. How are you? It's so exciting to be here.. 

Anthony: Thank you. Yeah, I'm good too. I know it can be a super busy time at Microsoft at the end of the year with, the half year reviews and things. Did you manage to get some time off over the holiday? 

Claudia: Oh yes, absolutely. In fact, it was just fantastic. We had snow here in Seattle, which is a rare thing. We're back to rain right now, but it was a fantastic time. How about you? 

Anthony: I'm in Southern California. But it was cold, so it actually felt like Christmas for the first time. So, I was going to say, I know Microsoft has such a broad range of products and services, and I should say I know that there are several billion-dollar businesses hidden within Microsoft.

So, what is specifically a Microsoft doing in the Agri-food and RCG space? Is it, different say to how they might approach finance? 

Claudia: Yeah, you're right. Very few people actually know that we are invested in industry as much but really in agriculture and food, it started about, I want to say seven years ago.

And it really started from a point where we wanted to understand the specific requirements and challenges that this industry is facing. Like data sharing, compliance, data privacy. But what we found is agriculture was just in the beginning of this digital transformation. And, since then, I think we've seen almost an explosion of applications developed from, anything from farm management, for food transparency or supply chain solutions.

And it's not the traditional it companies as much. And not only startups. But digital divisions of traditional ag companies, even insurances banks, or the large consumer brands or retailers that look into agronomic or solutions for the market. And we found that every organization starts from scratch and developing their own solutions.

So, we thought how we can enable data and analytics and insights better with a common data platform for the agri-food industry and anything beyond like cosmetics. And this platform then can help automating data sourcing data cleaning. And just organizing data in a way that you can run a two special models on it.

You can build agronomic solutions for any given area of interest. And we see that as a significant opportunity to accelerate the development in the Agri-food industry and being able to run solution on quality data pipelines. And we're doing similar things in other industries, but certainly. Agriculture, like I said, it's just this need for catching up on digital transformation is what we're seeing right now.

But how is this for SWARM? I know you are working with lots of those big companies out there in the market. What are you seeing in the Agri-food space on your side?

Anthony: Yeah, we're seeing many of the same things I'm sure as you, in that, when we first came into this, I have to say we didn't originally start in Agri-food.

We were looking simply to solve combinatorial math problems. But we actually stumbled across the first project that happened to be about food blending. And it was a great fit but then we solved that particular challenge. And what we really did was just uncover the next problem in the supply chain.

And we saw that as we explored this further, there were many challenges in supply chain, which we felt our technology was uniquely placed to solve. So, we ended up in the ag sector because it was specifically undergoing such a huge transformation, but we wanted to combine the supply chain issues we're seeing with the Agri-food sector. 

I actually like what you were just saying when you were describing the way Microsoft look at it, I love the fact that you've taken a real value chain approach and you're looking at the full end to end. So, you include retailers as well in that Agri-food supply chain.

And we've seen a huge number of challenging things happening in the market. Obviously, stuff like the alternative proteins, the ultimate foods and the plant-based drinks and so on, but also, we've seen new waves of data starting to appear, thanks to some of the IOT innovations.

 Information from fields or farm vehicles, or satellites and drone data. But I think for us, we ended up actually in Agri-food because we saw where it was a classic burning platform as people say. The existing firms have to change and many of their challenges were going to require a new approach.

So, we ended up here because it seemed like a market where we could make a really big difference. And I know you must have seen a similar pattern in the organizations you've been working with. What were the biggest challenges you've seen in the space so far? 

Claudia: Yeah, Anthony, I like what you're saying about burning platform and one thing you already alluded to is one when you look at farm to fork, the challenges are different. If you're an ag input provider, a food processor or your retail company at the end. And I think what's really the combining need in the industry is just being able to make better decisions based on data and analytics. And, we see the challenges are really ranging widely from, brand saying we need to be able to source more sustainable and be able to create this consumer transparency about how we source products, what went into growing it. There are things like consumers caring about animal welfare, about things being ethically grown, and you want to be able to bring this transparency, even what's in their food and what nutritional value is in there. And then we see the whole part that you also talked about Anthony, about improving operations and supply chain, understanding product availability and demand. We see tons of research, different research purposes, right?

I mean that comes from growing more resilient seeds. And also understand what new products are demanded in the market. So, I think we'll see a much higher level of personalization in products. We see a much higher level of product variations that come from regional differences or dietary restrictions or allergies or so all of that needs to be able to address.

And that ties back all of this ties back to technology. And if, if this is farm management for growers advising them, or is it about, being able to understand how a genomic composition of a seed reacts on the certain environmental conditions. I think I would be interested to hear from you because what we see is that there are a lot of challenges out there, and a lot of them get stuck in this initial where do I get data? Are you seeing similar issues? And are you seeing companies getting the full value out of what they're working on in terms of data and analytics today? 

Anthony: Yeah. That's a really interesting question. I think you painted a great picture there as well of the burning platform.

It really is it's not just burning it’s blazing. There's a lot of real challenge there and we are seeing a lot of the same things. Plus, some of those classic issues of supply and demand balancing, or network optimization. Where to open or consolidate plants or distribution centers.

But I think I'd also add in, the pandemic and we can't ignore the disruption that that's caused and some of the challenges there as well. So, I think handling disruption is one of the big things we've seen as well. And to your point on the data, we actually often see organizations have one of two scenarios.

They've either got so much data that they're overwhelmed with it, and they need help managing cleaning and making sense of that data. Or they've got some data, but they're missing. Key components to really make a decision. And I have to say the most commonly used tool for making decisions still in every business we talk to is excel it's.

It just is, it's just what happens and how it seems to work. 

Claudia: Not that I don't like that Anthony, right? The important point really that you're making here is even people expect to see certain. Answers when they look into data analytics, instead of being able to predict, for example, disruption, right?

Let the data telling you that there is going to be a disruption in the market. And it's just, it hurts me if you don't get the full value out of what you're investing in. 

Anthony: Exactly. I couldn't agree more. And I was going to say, if we drill down on a specific example and one that I know you mentioned, and a lot of people care very deeply about, and that's sustainability is that something that your customers are really talking about yet? Or do you think it's still just press hype? 

Claudia: No, not at all, I would almost say there is a sustainability angle to almost every conversation we have today. And when you think about climate change, the issue of food waste and impact on greenhouse emissions and water.

It's just natural to talk about those things. But I think what we're seeing is again, brands thinking about regenerative farming practices we see things like carbon modeling, practice verification, the. Growing need of, or the growing market of carbon markets and also things like being able to automate the measurements and KPIs you need for your ESG scorecard and things like scope three emission tracking.

So really broad in fact if I would try to. Capture it's, there is a measuring monitoring piece where you use remote sensing data. There is the modeling side of things where you on the, of the causal relation between like input management, environmental factors and where you were, where we then see this next step to simulation.

And what if analysis and then the last few pieces, the execution, like how can we reduce, or fix precision apply input. So, water as a third angle to it, but yeah, it's really broad, to be honest., 

Anthony: it is now those are interesting. And I like your definition of the different categories. Is there, are you seeing that the approach is a and the attitude I guess, is different from say a big retailer to a grower or someone working behind the farm gate?

Claudia: Yeah. And the thing is it is certainly different, but then in reality does all need to become holistic concepts. And I think what the big next step for the industry is to create an incentive for those regenerative practices on one hand, and really seeing the value in doing those things.

And then also making sure. Concepts can be done on a continuous programs are sustainable as well. So yeah, I think I'm definitely different, whatever part of the value chain needs to contribute, but it also needs to be able to go hand in hand. Then here we talk about being able to share data along the value chain which is an important component.

Anthony: Absolutely. And I think it's interesting, all that data sharing, I guess the first aspect is you mentioned before is the data collection. And I wonder, particularly for the growers, how hard it is for some of that data collection. Are you seeing challenges for say the retailers gathering the data from their suppliers or do you think that's been solved?

 No, it's not solved. And I think a lot of what we're seeing in the market today is that data collection is awfully manual and done by surveys. And I think there's a lot of areas of improvement and this is not meant to away the importance to collect data from farmers directly, but there's just areas that can be better seen by remote sensing, better judged by remote sensing.

Claudia: I give you just one example to be able to say, how much prep has been emerged or to see if a farm been tilled or not. I think that's all stuff. Actually, it can be done quite good over remote sensing and doesn't require like manual input. 

Anthony: Yeah, absolutely. Funnily enough I actually had a conversation with one of your other farm beats customers the other day earth sense, and I'll give them a call out here because I know a lot of people have been doing satellite work looking down on the crops and earth sense, have a little robot that goes under the canopy and looks from the other angle, which I thought was really interesting.

And I think they were one of the original launch partners for farm beats as well. I think that type of remote sensing from either above or below is really fascinating in terms of the type of data collection you can gather. 

Claudia: Yeah. And I think what's important to understand is its data consistency as well.

So, a lot of what we're talking about with the data platform we're developing is data fusion. So can you rely on. vastly available data if it's weather or satellite, and then improve the insight by having better local sensors, image, data and cameras are going to play a super significant role in the future for sure as well.

So yeah. Is this kind of what you're seeing as well, Anthony or, what is different in your conversation? 

Anthony: Yeah, it's very much, I think we're seeing the same things in the market there. And from my perspective, I've found most companies are looking at sustainability in four ways. There are the ones that have been in it for a long time and see it to do the right thing.

 They just are fully behind it. Whereas the, there are some that are doing it to just manage or enhance the brand. But in most cases, people are still looking to reduce costs. So, there's business benefit. It's not just a measure or a scorecard and a PR exercise. There's actually a business value in doing this, reducing water costs, for example.

And obviously there's also people doing this because of the regulatory demands that are coming through as well. But I think. What I have seen is that the growers and the bigger retail firms have quite a different perspective from each other. I actually spoke to one retailer who I felt remain nameless, but they switched to a different product group on the shelves for specifically one that had packaging that highlighted the sustainability aspects of the product. And they were telling us that product completely sold out very, very quickly. And they love this because obviously it's great for the brand. And it's also great for the bottom line, but at the same time I was talking to several growers who told me that in north America, in the UK, they're not happy with the extra red tape and the form filling that had been required by the same retailers because it adds a burden when they're already in a pretty tough situation. So, it's interesting. I do see a different perspective from the different parts of the supply chain. 

Claudia: Yeah. It's all about incentive though, right? I think that's really what it boils down to and for the brands, I think I can absolutely see it being preferred by consumers because of the transparency, because they care about sustainability.

Yeah, but much to do. Really, to be honest, I am not saying it's solved. There's just much to do to really make this more seamlessly, automated and attractive for the entire value chain. 

Anthony: Exactly. And I think the other point I'd make is that when you look at those ESG documents and the requirements, they're pretty large, it's actually quite a big scope in there.

And I have also spoken to several firms who have struggled with how to do all at once. So, it's the classic, operating on every part of the body at the same time. And so, most of the firms we've been talking to may have a score card that covers several areas. But really, they're typically picking a few areas to focus on like water and soil, for example.

But of course, carbon is always there because everyone seems to start there in that it's easier to measure it. And it's also very much in the public health. 

Claudia: Yeah, and I truly think it's important to look at this across the company, right? If you want to implement something trust from an ESG scorecard perspective, that's valuable.

Don't get me wrong, but what you're collecting as an insight might as well be super relevant to the research division in your organization. It's certainly interesting for your procurement team is interesting for even the, it can be interesting for the reason we just talked about for the marketing team to create more visibility about it.

So, what we see is working really well, if you not try to solve it just from one division but make it more of an investment that really pays up for the organization. 

 So, Anthony having said that, where do you see the companies are getting the most value out of their investment? 

Anthony: That's a great question. And I think we come at this from an optimization perspective. We have an optimization platform at SWARM that solves challenges for people.

And I think with sustainability, if it's just done as a score card, as you described earlier, that's more of a PR exercise. I think, where it really makes a value and gets embedded is where people are actually using sustainability to help them make decisions and better decisions in the organization.

And. Often when I look at sustainability, what it really is doing is adding another layer of constraints to decision, which can make it harder to make the best choice because of the rising complexity. So, what optimization can do is help balance those variables and make a choice that saves money and improves the metrics, lowering, carbon or water usage, for example.

So, imagine Claudia, you're buying a product like soybeans. And you've got growers all over the world and you need another 500 tons, and you went to get the best price, but you also need to think about those corporate metrics to lower, maybe synthetic fertilizers or water usage and carbon. So, some of those growers you're looking to buy from might have great stats on using organic fertilizers and they might have a low price point, but they could also be a long distance from the location where you need the delivery.

So, their carbon footprint is going to be higher. Whereas you might find there's another farmer with a higher purchase cost, but lower water usage stats and they reduce shipping costs and therefore carbon. So how do you balance all of that to make the best decisions? We think that's where AI optimization can really help.

And I was actually discussing this with a customer just before Christmas. And I think he mentioned that there'd been some confusion in his organization around sustainability and how to best approach it. But I actually think it boils down to five key questions and they're very similar to what you were describing.

It's number one, what sustainability metrics do you want to measure? Number two is do you have access to relevant and the right quality data for those metrics? The third one is what are the key areas you can directly impact and improve? Cause some of the ESG metrics are quite hard to actually tackle.

And then the fourth one is. One of the levers you can pull on in these areas. And this is something that optimization is great at. You can experiment with AI to change those levers and see what the best values are. So ultimately question number five is how do you optimize the levers to maximize business benefits and increase sustainability at the same time?

And I think often businesses see those as opposing challenges, but I actually think that they can be complimentary. 

Claudia: Yeah, that's such a great point. I love the questions and it really also talks to that. A lot of those constants get stuck in question number one or two. And you're not getting the value.

You're not getting the optimization out of it. If you don't think it through. And honestly, Before, I'm just a couple of minutes before it's this, if you start by the purpose of making better decisions based on data and analytics, so it's really the decision part and then work your way backwards.

You likely will come. And then you sort of those questions to design that. I think you; you are much more likely to get value out of your investment in, and analytics. 

Anthony: I completely agree. And I think, with so many people, including us, I have to say talking about, AI and analytics and in sustainability, I wonder how Microsoft looks at making AI more accessible and not just a specialized tool for experts.

Claudia: Yeah. It's such an important point, and accessible in itself means multiple things to me, at least accessible means, first of all, can I even have access to technology, right? And that's in, particularly when you look at the global footprint of food production and food security that becomes more relevant to start with, but then there's the other part around accessible in terms of, can I democratize AI, can I make it accessible to, people that are not data science experts, which I think is as important.

And then there's an ethics part, right? How can we protect the interests of everyone who is involved in terms of their privacy and data protection in terms of security? So, all of those play an important role. And I think in agriculture probably touch on a couple of examples. We have investments made in something that's called a planetary computer and a program that's called AI for earth, which really is our tools for the science community out there universities research organizations, and. How to stimulate sort of the innovation that's needed, especially in agriculture and food, which is highly regular, highly crop specifics, and really help organizations to work with those tools. And then we have the urban team. That's just, you're looking into connectivity and bringing technology, access to technology also to rural farming which is as important. So really, I think high priority and I'm just very glad we have the opportunity to do some things that are specific to agriculture. 

Anthony: Yeah, that's really interesting. I know Microsoft even from my time they were doing some great work in that.

And I know in some of the third world countries in particular, where they've made huge leaps of generations of technology, where people have got apps on their on their smartphones. They can do things that are just phenomenal compared to what was what was there in the previous generation that would have been something that they'd just completely bypassed.

I think the opportunities for actually helping growers with coffee, beans and cocoa beans and so on is immense. 

Claudia: Yeah, absolutely. We have seen a program successful that basically the nothing else, but I like sending text messages, 10 text message during the planting cycle and giving advice or our team in Africa is actually working on a bot called Kusa bot that.

To get a with partners, we're helping to get farming advice to farmers out there. Absolutely I think in particular is really a great opportunity to optimize rural farming with technology as well. But speaking more about the general notion, not needing to be a developer or a data scientist to work with data.

I know this is important to SWARM as well. And you talk about the no-code approach. How do you think about making a AI accessible? 

Anthony: Yes, it's a great question. I love your definition of the different aspects of accessibility too. And there was even the third one of course, with the visually impaired and those aspects as well, which we didn't touch on.

And so, we're always thinking about all of those different aspects, but when we talk about the no code side, we are very much believers in the fact that business users should be able to. Define a problem without having to do any coding or being, machine-learning expert, I have a degree in math and that the AI should solve it for them.

As you mentioned earlier, it's all about the democratization of AI. And I actually think similar thing happened in business intelligence and BI. My funny enough, my first job was actually at a firm in the UK, which became Kraft foods. It was general foods at the time I was there, and we made Maxwell house coffee amongst other things.

So, we had a lot of coffee beans, and we were grinding them up. And we had a very big facility. And once a week I would go to the warehouse and talk to the the warehouse team about what reports they needed. And I would then walk the mile and a half back through the facility and spend the next day or two writing reports in a specific language called east plus, which was a great report writing language.

And I'd produce these reports. And a few days later, they the team would actually have a report they could use. And you think now what would happen is they'd probably be using power BI or something that the people in the warehouse would make the selections themselves. And we'll be able to generate the report completely themselves without me having to go away and write anything and do any coding.

And so BI has become completely accessible for those end users in the warehouse. And I think the same will be true with a. I think ultimately, we don't need to have, people in warehouses or in the supply chain or in a truck needing to understand deep reinforcement learning. We, they know what their problem is.

They need to be able to just define that problem and then solve it. And so that's a major thrust behind what we're doing at SWARM. How can we make it easy for people to define a problem and just let AI solve it? That is what we're doing to really make it more accessible for everyone. 

Claudia: I like it, Anthony, you just revealed what your motivation is, for completely understand. 

Anthony: Claudia, where we're at the start of a new year how do you think the market's going to change in 2022?

Claudia: Yeah, very good point. And you already mentioned COVID before. I just think we are going to see persistent challenges because of COVID. We already experienced that right now in terms of supply chain. Issues in the market we, that contributes to for the food waste. That's just a side effect of what we're seeing right now.

And I also think the labor challenge is going to drive the need for more automation and robotics in the future. I think we've seen tremendous examples and it feels not tremendous, but actually painful examples for fields not getting harvest because of the lack of labor in the last year.

 But what we see is I think with consumers being more involved in their own food preparation during COVID this need for consumer transparency, what we were talking about before is a really understanding where does food come from, is it that, is it nutritional valuable for me? For my personal health benefits? I think we'll see more emphasis in that sense. And I think there is a new trend coming around, more personalized nutrition, and I think it's going to raise a whole bunch of new requirements, both in supply chain planning and, food innovation in the industry. So, I think those are trends that will impact us from a technology perspective. For sure. 

Anthony: That's yeah, those are really interesting points. And I have to say I'm in complete agreement there. K myself about what I see happening over the next 12 months. And then I come to the same conclusions around some of those areas.

I think I would add food safety and labeling issues, particularly with that. Some of those that those changes in the way people are looking to consume food. I think food safety and labeling will become a bigger issue this year. 

Claudia: And I hope it can be done better. Yeah, you have to stand around food safety today is really around audit data.

And there's so many ways how technology can predict issues in that process and the food value chain. And I, coincidentally Anthony, I looked this up yesterday, but a number of food recalls actually increased again. There's certainly a lot of room for 

improvement. 

Anthony: Yeah, that's interesting and I, I can't reveal this. I will say we got a set of data this morning about a certain type of animal protein and e. coli. We're analyzing that to spot the patterns in that for a particular customer. So, we're seeing actually people coming to us, asking for projects in that area specifically. So, I think that will grow this year as well.

But I think obviously the sustainability we've talked about will continue to grow in importance with budget as well. I'm seeing more people setting more money aside for that. But I think the other one, and clearly you mentioned it labor. I think labor shortages and disruption will continue to drive innovation in both robotics and AI.

But I think labor shortage is a really interesting one because. It affects both ends of the market. You've got the agricultural workers picking fruits and berries. The, as you said, fields have been left on harvested, but at the same time, you've got a lack of highly specialized data scientists that to, to support some of the Agri-food initiatives that these companies would like to make.

That's why we believe it's so important to democratize AI, as you're saying our mission, make it accessible because that. There's huge wastage still in the food system. And there's a great opportunity for improving all of that for everyone. 

Claudia: No, that's really a good point.

And it just reemphasizes what we were just talking about, making AI accessible to everyone but maybe switching topics a little bit here You want to talk a little bit, Anthony, about the partnership between SWARM and Microsoft and where we're SWARM is on the Microsoft partnership spectrum.

Anthony: Absolutely. So, we're clearly quite closely tied because our C level team all spent at least five years at Microsoft. I wouldn't say tied at the pit, but we know the organization very well, and I've got a lot of contacts and a respect for what the company does in this space. So, we've been going through the standard partnership program right now we're in the Azure marketplace already.

We're also Cosell ready, which as Claudia means the Microsoft field sales teams can part with us. And get credit against their sales targets for joint sales. But we also have several team members going through the Microsoft certification process so that we should be silver partner status later this year as well.

And all of those things, really what that means is we can get more support in the field to actually be able to support the customers and make sure that there's a great experience for them. We're really pleased to be moving through that process and getting further on with that. 

Claudia: So that's a, that's great news, Anthony and I can only agree that it's important to understand Microsoft and with the experience you have at really helps navigate. So, I do think there's a really great opportunity between what we're doing in the Agri-food space with this platform we're building, and you're focused on data and analytics and being able to help organizations that are building out their digital strategy to develop their solutions on the platform. I don't know if you would agree. 

Anthony: Yeah, completely. And I know your team and ours will continue to work on this. I think there's a great opportunity to help people get that end-to-end value chain optimized as well, and get the underlying analytics in there, too.

Claudia: So, for people that are joining us today, what would be the best way for them to engage with SWARM and your team in the market? 

Anthony: We're always looking for interesting projects, especially where we can help optimize key processes, whether that's, balancing supply and demand or managing disruption or allowing the sustainability metrics into the existing decisions like we talked about earlier.

So, people can reach out to us via our website or drop me an email directly at Anthony@swarm.engineering. Um, But I think I think that the key is we're going to continue to work on a whole variety of different areas over the course of the next year. We'd love to partner with some other organizations in doing that.

And clearly, we're linked with Microsoft very strongly and I'm looking forward to seeing the fruits of your work, Claudia over the course of this year. And I just want to thank you for joining me on this first SWARM fireside chat. I'm hoping we can do it in person next time, but yeah, looking forward to an exciting 20, 22, working with you and Microsoft.

Claudia: Thank you so much, Anthony. That was absolutely fun and enlightening on a rainy day here in Seattle. So, I truly enjoyed the time and uh, thanks again, for um, inviting me. 

Anthony: Thank you. 

Holly: That concludes our first episode of the SWARM engineering podcast. Thanks for listening. And if you were looking for us or would like to send us a message, you can find us on LinkedIn, our contact links, as well as others mentioned during the show will be listed in the show notes on our site, www.swarm.engineering, and make sure you subscribe to the podcast. We have some really great guests lined up for upcoming shows that you really will not want to miss. And one last thing we need help with the name for our podcast. Can you help us come up with a winning name? We really would love your help.

You can go to our website to cast your vote or make a suggestion at swarm.engineering/nameourpodcast. Thanks again and see you next time.