Sramana Mitra: Were you still based in Georgia?
Radhika Subramanian: All this was happening in Atlanta, yes. Literally, on Georgia Tech campus. I launched the company at ATDC which is the Georgia Tech Incubator. They reached out and funneled millions of dollars into Emcien. Then, we started to do the first implementation and the payback was huge. It was phenomenal. The customers were Caterpillar and John Deere.
The dealers of these companies are huge. They are big billion dollar companies. We did a private placement with Adco, which is a dealer of Caterpillar, and did our Series A round. We started to build the company. Then the dot com burst happened. When the manufacturing and financial downturn happened, I thought, “Okay, Radhika how are you going to get up from that?”
While this was going on, I was also starting to get a feel that manufacturing is ripe for change. There’s so much potential in manufacturing but the manufacturing sector is always going to be slow to adapt. Big Data and data analytics is not something they understand. We put together this group and they advised me, “Your engine works really fast. You have this fast pattern detection engine.”
Now let me take a step back. This is really important. Manufacturing data is highly unstructured. It’s called deep data. When you take a tractor and engine, you’ll get about 600,000 number of features. It’s very deep. Nothing can really analyze that data. The analytics and the pattern discovery that we have created were like magic on that data.
Sramana Mitra: What year are we talking about?
Radhika Subramanian: Around 2009. We took the engine and started expanding it out to different types of data. We can crunch retail data in real time. It’s very fast. Then, we extended it out to data from telecommunications.
I know manufacturing like the back of my hand. I had an amazing marketing group. In April 2013, we decided to host Big Data Week in Atlanta. We thought, “This will be a great way to sponsor the event and also gauge the maturity of the market.” We can at least find out through conversation and figure out what solutions do exist. How do people talk about it? What are they looking for? Long story short, I understood the market maturity.
They told me not to say that the market is not very early. They’re looking for solutions. We already had some of the key things that they were looking for. That increased the pace and I said, “We’ve key parts of what’s required – automation being a big one. We just need to start scaling and start moving across the sector.” We ended 2013 negotiating a reseller agreement with MCR for retail. We are now expanding into a few more sectors.
The one thing you need with data is economics of analysis. We need to get to the data fast. It’s mandatory. It’s not a luxury anymore.