Rosslyn is home to countless influential and interesting businesses. In this series, Business Buzz, we chat with the leaders of these companies in our community. These are the thinkers and doers who make Rosslyn tick! Want us to feature your business? Contact us!
Founded in 2015, Fresh Gravity is a business and technology consulting firm that helps clients achieve their goals by adopting transformative technologies. They specialize in data management, data science and analytics, artificial intelligence and application programming interfaces (APIs). Their corporate headquarters is in Santa Clara, Calif., and they recently opened an office at Rosslyn's Eastern Foundry campus. They also have offices in India and Australia.
We sat down with Fresh Gravity's Co-founder and Co-CEO Nilesh Dhingra, Human Resources Manager Rebecca Graff and Capability Lead for Artificial Intelligence and Data Science & Analytics James Ray to talk about the company's work and future direction. Below is an edited transcript of our conversation.
Rosslynva.org: You do some really interesting work. Curious how the company got started.
Nilesh Dhingra: Fresh Gravity was started by me and Ajit Kumbhare. We've known each other for many, many years. Ajit was a successful entrepreneur who had started another company in the Bay area. I've been in consulting for 20+ years. We wanted to focus on some cool technologies that can be applied to digital transformations while building an organization that has a unique, friendly and open culture. We don't have any investors. We started with clients we found through our network initially. Since then, we've done work for clients like Intel, Symantec, Pandora and LinkedIn.
Rosslynva.org: What are your main focus areas for 2018?
ND: First, we will continue to focus on data management. Today, clients have more data coming in than ever before. For instance, social media results in significant volumes of data. And entities such as utilities now have large volumes of meter data coming in every few minutes as opposed to a monthly read. How do you manage this data? How do you structure it? How do you analyze it? It's said that only 3-4 percent of data is actually used. The rest of it is just sitting there, and it's a gold mine because there's so much that can be done with it.
Rosslynva.org: Plus, you need the capacity to store data. You can collect and store it all now, and it's really important how you structure the storage so you can actually use it.
ND: Yes, exactly! Our other area of focus for 2018 is data science and analytics, which ties closely to artificial intelligence (AI). We also focus on application programming interfaces (APIs). APIs are traditionally how systems were integrated. But now, if you are using an app on your phone, there's most likely an API behind it that connects you to the data on the back end. We help clients build APIs and secure APIs because security is a big concern.
Rosslynva.org: So you're not product developers, then?
ND: At our core, we are a consulting firm, although under Gravity Labs, our research and innovation hub, we're building niche products. To summarize: We take clients' problems and work them into technology projects that deliver clear, distinct, value-driven solutions mostly via tailored consulting and partly by offering high-end, product-like solutions.
Rosslynva.org: How big is your team?
ND: We have 30 staff in the U.S.: about 15 are aligned to Santa Clara and 15 to our Rosslyn office. We have 50 people in India and 20 in Australia between Sydney and Melbourne. We're all set to grow in 2018. We have a plan and a strategy and are looking to go for it.
Rosslynva.org: What do you think is most interesting about your work?
ND: We're doing a lot of great work in data science and AI. I'll let James [Ray] talk about that since he's our lead in that area. Our approach is different from some of the large companies like Google and IBM. We really understand the clients' problems and solve them in a very systematic, consultative way.
James Ray: Many other companies have an extreme focus on the technology and the specific data and architecture needed to produce it before they have a good understanding of the actual business objective. I come from a background that spans both corporate strategy and analytics. What I took from the strategy side of my experiences was this very disciplined design philosophy. I start from the end almost radically. I ask myself, 'What is the decision a client needs to make based on the data?' and 'What are the analytic results that would inform that decision?' We constantly interrogate our approach and speak with our clients to ensure we're moving in the right direction. That's what sets up apart: the discipline of our design philosophy, and our business savvy and careful listening. We take a consultative approach.
Rosslynva.org: Can you think of any specific examples where your approach was really effective for a client?
JR: Well, I could tell you about a very unique local client. We're running a data management and data science project for them. They maintain and manage microorganisms like bacteria, viruses, strains of Ebola, Zika and SARS. They research, reproduce and sell these microorganisms to multiple clients like the National Institutes of Health. As you might imagine, the processes they use to manufacture and pack and ship these microorganisms and associated accessories are very complicated and risky. Essentially, they are sending anthrax and Zika through FedEx! It's important to get it right: the kind of container you use; what types of materials can go in a container; what products can be packed together and what ones can't. Today, all of this is determined by human experts going through a shelf full of binders with all of the rules for packing these appropriately and legally. This results in a lot of inefficiencies and costs. Of course, if you have an order of 500 microorganisms, you can save on handling and freight if you can ship them in fewer boxes.
The project we're doing for this client has two pieces: First, we applied machine learning to learn the rules, so instead of having a human reading binders and piecing it together, we looked at past shipments and trained the machine-learning algorithm to infer what the rules were. Actually, we trained multiple algorithms. Not every combination that's possible had been done. So, that was the first piece: Train the algorithm to learn the rules. The second piece was feeding those rules into a second system that optimizes the packing. So, if you're the packing staff on the floor you plug in an order and you get out the system's recommendation of the packed boxes. It will say: 'Box type A with products 123, then certain quantities.' And it's optimized. Maybe before they'd have shipped it in 40 boxes, but now they can ship it in 20 boxes.
Rosslynva.org: That's really interesting, and you explain it so well. Are there other types of organizations that would benefit from your expertise?
JR: The short answer is that every industry will be transformed thought AI, but for some it will happen earlier than others. Processes ripe for automation share certain properties, and some industries are more ahead than others. We're seeing that the competitive gap between the organizations of high analytic maturity and low analytic maturity continues to widen. The companies that have brought in AI expertise are starting to grow much faster than those that have not.
ND: Any company that has data where you can automate a process and use machine learning is a use case. So, in other words, anyone and everyone can benefit from machine learning. That's where the world is going.
Rosslynva.org: I see you also work with county and local governments, even for improving snow removal and everyday things like that.
JR: Yes, we use something called time series analysis to understand patterns in emergencies that may apply to future events. For instance, we look at several years of open data from a city, which is broken down by location, zip code and even block. We analyze the types of calls that come in over time so we can say, 'Ok, next year on a given day in that block, we'll probably have a certain type of call.' That can have a cascading effect on lots of stakeholders. Ambulance companies can adjust their staffing and fire departments can do the same thing. We also have the GIS mapping expertise to create interactive dashboards for these service providers so they can adjust where they place their staff on a given day. So, someone using our dashboard can come in Monday morning, fire up their laptop, and immediately see the first five things they need to do.
It's a pretty wide leap to go from what is predictive to what is prescriptive. The models that tell you what's going to happen are not always the models that tell you what to do. That's where AI and robotic process automation (RPA) skill comes in. They're tuned to not just what's going to happen but what's going to happen and what you can you do about it. We offer that, and it sets us apart.
Rosslynva.org: How do you like being at Eastern Foundry?
Rebecca Graff: It's a great space. A very collaborative work environment. Very open. Everyone works with one another, so it's been very positive.
ND: They have a lot of events that we take advantage of, including technical events. It's not like other co-working spaces out there because it offers not just co-working but also networking. We've definitely made connections at Eastern Foundry with companies we may collaborate with in the future.
Rosslynva.org: So, are you hiring?
RG: Yes! We value certain qualities in employees: passion, integrity, openness, respect, adaptability. . . . I see my colleagues exhibit those qualities on a daily basis, and we're looking for people who also share those values. [See job openings at Fresh Gravity here.]
ND: Rebecca is the keeper of who's coming to the company, which is a very important role. The quality of our people is one of our biggest differentiators: the talent we already have and the talent we want to attract. We have a truly open and respectful culture, and that's also very important to us.
More info: Fresh Gravity