FROM MAGAZINE: Interview with Nishith Rastogi, founder of Locus

Nishith Rastogi worked with Amazon building algorithms to counter credit card fraud. He co-founded RideSafe, a women safety initiative that uses cutting edge algorithms to provide a safe commuting experience. Rastogi holds a graduate degree in Electronics and a Master's in Economics from Birla Institute of Technology and Science, Pilani (BITS Pilani). He has published papers in the field of experimental physics and has to his credit pending patents in machine learning.

In May 2015 Rastogi founded Locus. It is a technology platform for optimising logistics operations using proprietary algorithms, powered by data analytics tools. Locus helps enterprises reduce logistics costs with increased efficiency, on-time deliveries and a delightful end-customer experience. Rastogi’s team includes engineers from Indian Institute of Technology and BITS Pilani. He has data scientists with PhDs from Carnegie Mellon University and Tata Institute of Fundamental Research building the next generation of logistics industry at Locus. The company is backed by academia, domain experts, ex-entrepreneurs and great investors. Nishith Rastogi, founder of Locus, speaks to Reji John about loading intelligence into the complexities of logistics using modern science and advanced data algorithms.

What are some of the favourable conditions for the logistics startups in India?
The primary factor that I would call out is the amount of change that is happening right now. Every industry has a very macro factor when it comes to disruption. Think of a shopkeeper who orders an Uber and it comes in five minutes; he orders something on Flipkart and it comes on the next day. Now he is wondering why is Britannia not able to replenish my goods the next day. So there is a huge amount of change that the industry demanded which effectively made logistics a game of data. That is where a lot of startups came up.

What are the unfavourable conditions?
Data. For example, when I do routing in Singapore one simple thing is how to convert an address to a Lat long, which is the sole problem over there. But in India I need to develop almost a entire system spending more than a year; a proprietary geocoding system which can understand the way India writes address. Lack of structured data is the biggest impediment here.

How do you intend to change that?
You defeat it with technology. It is just that you need to build it with far more advanced level.

To what extent technology disruption is guided by consumer behavior?
Definitely to a large extent, even though consumer demand in itself is just 10 percent of the logistics market. The impact of the consumer demand also goes directly to the B2B distribution book. Previously B2B distribution was a 7-day replenishment cycle; today it is a 48-hour replenishment cycle. What B2C logistics has done in a very fundamental way is that previously the logistics to a consumer was a pull model. Consumer used to walk into a store and pull from there whatever he wanted. But in the last few years, the state of B2C logistics is evolved to a form where it is a push model; whatever is needed to a consumer is delivered into his or her doorstep.

Speed is essential for any business to run successfully. That’s where the technology comes in. To what extent is your solution built to plug in to such a scenario?
It is very well designed for that scenario. There is something that is known as network effects of automation. If there are three things like ‘A, B and C’, and if A and C get automated, then when B gets automated, it improves the performance of A and C. Then the systems can talk easily to each other. The second most important aspect of automation along with efficiency is also consistency. When you automate something, you can expect a very predictable behavior. In large corporations, it’s even more valuable than efficiency.

How do you bring in solutions that break up the complexity of logistics and offer a robust system?
Logistics is effectively solving a graph. You have things moving from certain corners along certain edges which have different properties. Solving a graph is a kind of mathematical problem that is very natively suited for computers. It is much like a stock market where there are lots of moving variables at the same time and one needs to figure out an optimum short term trading strategy. The moment stock markets got connected, quantitative models took over technical trading. Fundamental investing, which is long term, still remains with humans because humans are good at pattern recognitions. Computers can solve the problem that exists in the nature of logistics. When computers don’t solve it, what happens is either you are simplifying the business problem or you are producing unoptimised solutions. By unoptimisation, you are using more number of trucks than required, you are using more resources as well as more fuel. Or by simplifying the business problem, means you are not doing an optimal number of dispatches.

How receptive are your potential clients to solutions built on big data, machine learning and airtificial intelligence?
One of the things that works well for us is our proof of value which is very demonstrative. What we do is we take one week past data of the company and administer into our systems to optimise the routes for the client. The client can check whether the solution helps them to save money.

What are your service offerings for the industry and who are your clients?
We work with companies like Big Basket, Urban Ladder, LensKart and some of the large third party logistics providers in India. The key problem we solve for them on an everyday basis is that if a company has hundred packages to be dispatched; How many vehicles should I be using to dispatch? What should be the size of those vehicles? What group of packages will go in each of these vehicles? What route each vehicle should take to deliver them in specific chronological order.

What is your vision for Locus?
We want to create technology that will automate all the human decisions that are required to transport a package. We are a team of 45 people. Our company is less than three years old. We are operationally profitable and we do not know when we will break even. For us, it’s a very product driven sales.

As a startup, what is the biggest challenge for you right now?
The biggest challenge right now is to get the best and the right talent from all over the world.

What’s the impact of Goods and Services Tax (GST)?
In India, when you set up a warehouse, more than being a function of logistics, it was a function of taxation. Now with the introduction of GST, companies are optimising their warehouse locations for logistics. And that creates opportunities for players like us.