It's a dream scenario: choosing your own cloud platform when designing, architecting and building a global cloud enterprise software application. And that's just where the Fulfilld story begins.
Fresh off the launch circuit the SaaS company is breaking the fourth wall and is taking the C2C Google Cloud customer community behind the scenes and along for the ride with their development, engineering to business leadership teams.
They'll candidly share their successes, challenges and your engagement is welcome. This series will be a mix of articles, discussions, on-demand content and even live events where you can bring your questions and comments directly to the teams.
To kick off the journey, we begin with understanding who FulFilld is, why they chose to build on Google Cloud and how micro-services are enabling them to quickly deploy features, develop an intelligent enterprise warehouse management platform and support high-volume transactions that can scale globally.
First Things First - Why did you choose Google Cloud?
Michael Pytel, CTO, shared that he and his team is working to deliver an enterprise-grade application that enables a warehouse digital twin using 5G ultra-wide-band powered devices.
From supporting high-volume transactions across the globe, to analytics, to machine learning and natural language processing that powers an industry first warehouse digital assistant, Google Cloud Platform became their go-to platform when looking at functionality, pricing, scalability, performance, and innovation.
Listen and Join the Journey
In our first conversation, Pytel and C2C cover the following:
- What key decisions contributed to choosing a cloud platform for the SaaS application
- FullFilld's requirements for a globally available application
- Why they need a combination of in-memory databases (Firebase) and traditional SQL-based database (Cloud SQL)
- Why they were so focused on leveraging the autoscaling features of Kubernetes for application logic
Rather skim?
Key questions are shared below along with the full transcript with edits only for clarity.
__
Michael Pytel, CTO, FullFilld (MP): Fantastic, Sabina. Thank you so much. I've worked in enterprise applications, really all my adult life. So I started as a night operator supporting an earpiece system called pix. Then I supported JD Edwards and PeopleSoft and then SAP, enterprise ERP and spent the last decade there.
Now with FullFilld, we're building a brand new company, and at this brand new company, we create the digital twin, which is really just a digital representation of the physical warehouse...so that you can visually look at how do people move in my warehouse, and how inventory moves in my warehouse, that's our secret sauce. That's our thesis as to why we're going to be successful and we're getting a lot of good feedback for the market on our product today.
Sabina, C2C (C2C): So tell me a little bit about that feedback, what's resonating?
MP: This is where Google Cloud Platform comes into play, a lot of our customers love that it's very low total cost of ownership. You know, there's no server to deploy on premise. Everything is cellular connected, and Wi Fi connected. So we have that backup, if the customer's Wi Fi network in the warehouse goes down, it falls over to 5G connectivity.
So it's a really low cost of ownership, really easy and quick to deploy and our application is auto scaling, which I think is another benefit of Google Cloud Platform, meaning our customers don't need to worry about running out of resources, right? As they grow from a 50-person warehouse to 100-person warehouse and then add the fourth, fifth, seventh and 10th warehouse, the application auto scales on Google Cloud Platform with the customers growth. So they never really have to worry about "am I maxing out the server? Are we over utilized? Is the system going to be slow when I add this new product line?
We don't need to think about this. We can think about the business challenges we have. We don't need to think about server capacity, which I think is a big benefit with running on Google Cloud Platform.
C2C: Yeah, yeah. So talk to me a little bit about that decision, then to go with Google Cloud. That is, did you build knowing that you would use Google Cloud? Or was that something that came up later? Talk to me about that decision.
MP: Yeah, so there's multiple, you know, infrastructure as a service organizations out there, Google, Microsoft, Amazon, you have other providers out there as well, that are smaller, but still very innovative. So we had to find that right mix of brick and mortar stability and investing in new technology, and constantly innovating.
When you take a look at Google and their ecosystem, the way that they share knowledge, the way that they share their product roadmap, the way that they create content on YouTube, for developers to watch and articles for us to read. We just felt like, "wow, this is a fantastic organization that's continually innovating, continually pushing the envelope of what they can do and what they can't do. Also, enabling startups like us to run an enterprise level application and enterprise grade application at the lowest cost possible [is another benefit].
So, it's cost optimized, super innovative, great content, great partner program, easy for us to learn and ramp up resources. That was part of my personal scorecard when determining what platform to run on. Google Cloud Platform really just checked all those boxes for us.
C2C: What are you most excited about with Google Cloud?
MP: So personally, you know, their ability to run essentially functions with it for your application and in an auto scaling manner. A lot of cloud providers can run Docker and Kubernetes, so Google has the Google Kubernetes engine and we can run code in the backend on Google Kubernetes engine, and it auto scales. Microsoft can do that, too. And Amazon could do that, too. But then there's this feature in function called Cloud Functions, which even further drives our costs of operating even lower. And they're really innovative. I can use TypeScript to node and Cloud Functions. This is probably getting, you know, super technical...
C2C: ...Our community loves technical, go for it, give us the details!
MP: Fantastic. So, you know, looking at Cloud Functions, we just loved the way that they worked in the function.
We loved how they're cost optimized and when users are logging in and using the application, it can auto-scale and grow. I don't need to really take on the management of Kubernetes and Kubernetes clusters and the management of how many nodes are active, I can just use this thing called Cloud Functions.
Another thing within Google Cloud platform that we really loved was Firebase.
We use a mobile application in the warehouse, where you can picture yourself as a warehouse worker, and a garage door is open, and there's a truck that you have to unload. A lot of times, you're not unloading that truck, and it has pallets and has different products on it and those products are going to different places, you don't typically unload a truck by yourself, you have a team member and a teammate or a group of people that are going to help unload this truck. So we needed a mobile application that was super responsive, super fast.
[For example], I receive 10 baseball gloves that are sitting right in front of me and somebody else grabs another 10 baseball gloves, we need to update each other, letting each other know that we both receive 10, we need to let the backend system know we've got 20 total, and we need that to happen very fast. So we're using Firebase, the cloud-based no SQL data in memory database, we're using Google Flutter to build our mobile application handling authentication there as well. It's just a very responsive, very fast application because of these cloud technologies.
You know, we could have gone a traditional API route with a traditional SQL database, but Firebase has been super responsive for the mobile application making it so that the warehouse worker can just keep working, keep working, keep working and not wait on the application to update the screen. So it's been fantastic so far.
C2C: How does that translate into the business outcomes for your customers or your clients?
MP: So within the supply chain world, there are tremendous pressures to get products to customers, right, we learned we read about this all through COVID. There's more DTC shipping and more direct-to-consumer.
Brawney paper towels, Georgia Pacific, they were so used to shipping whole pallet loads to Sam's and Costcos. Now they have to ship individual piece products directly to consumers. This is happening across the industry, so there's just more things, more movements, more activities, more documents in the warehouse. Anything we can do to make sure that the user in the warehouse is supported is not encumbered or you know, they don't view the system as a bottleneck, they view it as an enabler, that's what makes us look great makes the warehouse worker feel good about their job, meaning they know what to do, they know what they need to do and they can get it done very quickly and they're waiting for them on the system to process it.
So us being very responsive and very quick enables that warehouse worker to do their job effectively throughout the day and enables that organization to do more.That's what we're trying to do is we're trying to make it so the individual warehouse worker can improve their throughput by 50% by navigating them through the warehouse very effectively using Google's Auto ML and machine learning models for routing in the warehouse.
Being super responsive, supporting that worker throughout the day, and just enabling organizations to do more. That's our goal.
C2C: [Are you ready to compete with Amazon?]
MP: We get this question a lot. Amazon's a massive company and they run a lot of their own software naturally being one of the largest companies on the planet. In North America, there are 40,000 other customers and 40,000 other companies that make a product and need to ship their product to a customer. Our goal is to democratize the technology typically used by large companies and make it available to midsize companies.
So being able to create that digital twin of the warehouse, yeah, Amazon already did that. But they have billions and billions of dollars. You know, what about the $200 million manufacturer of equipment in Durango, Colorado or the, you know, the upstart shoe manufacturer, and they're making shoes in the US, and they're selling it directly to consumers? They want really sophisticated warehouse management application, something that's going to help them be super effective as they move product through the warehouse, reduce collisions, reduce the number of times we touch a product, optimize the picking route, the way that people are walking through there, what if they want that technology and they're not a billion dollar company?
What do they do? Well, that's where FullFilld comes in, really trying to democratize that large enterprise level of features and bring it down market into those midsize companies.
C2C: It's.. It's really cool...and... that's your that's your why, right? That's your big mission, your core every day. Was this a COVID born decision? Or where did this come from?
MP: We definitely founded the company in 2020. We started the company during the pandemic. We saw the need there. But there were also some other cool things that were in the mix here from a technology perspective.
There's a technology called ultra wideband, which is not specific to Google or anybody else, but ultra wideband is a location indoor positioning technology that enables us to understand where an object is in a physical space.
So there was a convergence of the need, meaning the need was COVID and direct consumer was going to continue to grow and that every analyst agrees, it's going to continue to get even bigger. So we knew that logistics and supply chain space was going to have growing pains. We had this new technology that's being adopted more and more and then we have Google Cloud Platform, which enabled us to stitch it all together.
Now we're using machine learning in Google Cloud Platform, to make recommendations to customers on where to store products in their warehouse. Because of this location technology, we understand where the product is and because of the application we built, we understand what needs to be moved, what are the orders, what are the sales orders, what needs to be moved to the customer. So blending it all together, they're on Google, it's really cool.
C2C: [How did the application begin?]
MP: When we started the application, we started with a UX design and beta test with a few customers. We created a website, we created the design of the application, we communicated to the market, what we were doing and what we're building.
As we were just in the beginning, just starting to build the application, we got pinged from Deloitte in Europe, and they had found us and one of their large retail customers had taken an interest. They said, "Wow, you guys are doing location tracking indoors to make employees more effective in the way that they move inside of the warehouse?
They thought, well, "could I use this in a retail scenario where I've got large retail? I need to move people around at night, because I want to turn my retail stores into warehouses
So while we have to be able to run in multiple data centers around the globe, that obviously, is something Google's very good at. We needed different features and functions within Google available in European data centers right away, which was fantastic.
Google already had a partner that also can help us understand some privacy laws in different countries and Google has a lot of information that showed us, you know, or gave us leads on how to handle different privacy and GDPR compliance within European data centers, which was fantastic.
So that was number one, we knew we had to run in multiple data centers, we needed to be multilingual. And again, just tacking on all the little components we needed. We needed an in memory database, we needed an attritional database, we needed to be auto scaling because we wanted to have, you know, really low operational costs, runtime costs.
The next thing that we wanted to do was build a world's first natural language digital assistant. So the Siri or Alexa, the Google assistant of the warehouse, meaning I could hold up a device, my Google Pixel, and I could say, "where's my next pick? Where's this material? What's the status of the next delivery? How many more tasks do I have? Natural language digital assistant on devices in the warehouse specific to my job function and Google offered that as well, the ability to have a natural language digital assistant in multiple languages. So we were able to use their application to build a digital assistant that can speak Spanish and Swedish and English.
As we continue to grow, we can continue to add more languages and be more global. So we definitely knew from the beginning, we wanted to be a global company, and GCP has those features to help us do that.
C2C: That's, that's awesome. Thank you for sharing that whole story. That's a nice succinct way of how it started and where you are now to set everybody up.
One of the other things that we offer to set up the community with is a solid criteria that can be used when determining the right platform for your application.
Are there certain key makers or decision points that you can share for others that are evaluating whether or not they want to build on GCP?
MP: Yeah, that's a good question. The big thing is the developer community and the developer community support, right? If you're transitioning a developer into a platform, are they going to be able to ramp up quickly on the knowledge required? Are they gonna be able to participate? Are they going to be able to have test environments and demo environments, at a very low cost?
So that one thing was just developer community and developer community support. I think Google is very developer friendly and supports developers.
The next one was service availability and data center availability.
Can I run in all the countries that I need to run? And Google had that check mark there in terms of innovation, you know, as an organization they have a well thought out roadmap. They clearly communicate to the community what they're building and what they're sunsetting. We need to know, as we're building an application. If we're using a specific technology, what does the roadmap for that technology look like within your company? Is this something you're going to continue forward with? Or is this something you're going to kill and create something new?
So as we are betting our futures on different technologies, whether that's Kubernetes or natural language processing, or x, what does that product roadmap look like so I can lay out my product roadmap. I think that Google's doing and has done a very good job of laying out what the roadmaps are in specific use areas. There's always room for improvement, we always want more information, right? I'm never gonna be happy.
But that's one thing you need to look at when choosing a cloud provider is, "what does the product roadmaps look like? How far out are they forecasting? And are they meeting the goals that they're setting so that you can plan your product around that company's product roadmap?
So product roadmap, developer support, developer adoption, service availability, data center availability, were kind of our top three.
C2C: Yeah, thank you so much. Is there anything else that you wanted to ensure people understood about the FullFilld or why you chose Google Cloud? Because my last question, then, if there isn't anything, which I'm sure there probably is, is why you are excited about the C2C community and how you see yourself contributing or being a part of the community?
MP: In terms of FullFilld, we want to share not only our product and what our product is, what our vision is with the warehouse management logistics community, we are also eager to share how we're building this platform with the development community.
We're eager to share our experiences, talk about it out loud, get feedback, and ask good questions. One thing I've learned in my technical career is the more I share, the more I learn and so we are here fulfilled, we definitely want to share everything that we're doing.
We want to share how we're building it, where we're building it, our timelines and the functionality that we're using. We're looking forward to engaging with the C2C community to have those open conversations because we're going to learn something that we didn't know.
Someone is going to ask a question of, "why did you do that? We need to defend it or adapt and move to something else and in. We don't want to build our application in a silo. There's a wonderful community of people out there and specifically within C2C and we want to tap into that community, solicit feedback, solicit ideas and hopefully find some people that want to work for FullFilld in the future.
Also, I hope that we are sharing enough information so that as other individuals are out starting their company, building a new platform or building a new application within a larger organization, they can learn from our mistakes, hear about our challenges, and adapt and grow from there. That's, that's what that's really one answer for both questions.
C2C: Yeah, that's, that's amazing. Awesome. That's all I got. Do you have other things you wanted to add?
MP: No, thank you so much for the opportunity, support, excited to share and really hope that we get lots of great q&a and questions from the community.
C2C: Yeah, me too. Me too. I'm really excited to share this out. And so thank you so much for your time, Michael, and I'm sure we'll talk with you soon.
MP: Yep, see you soon.