Cloud Hosting

Highlights

Magicbricks, the leading real estate firm, has chosen to grow its data science scale and is searching for individuals to join its data science stack for the completion of future foals. They are looking to employ more full-stack, mobile, and data science engineers as they transition into a full-stack service provider.

"We have expanded our data science team by 50% as we plan to utilize data science more in our work, unlike last year. By the end of this year, we wish to complete our team with 20+ team members, including AI engineers, data scientists, software engineers, big data engineers, and others,"Subodh Kumar, CTO of Magicbricks.

"This was not easy because tons of analysis and lots of insights were coming, which was leading us to data paralysis. So, we decided to restrict ourselves and started to focus on data classification and data moderation. We figured that other statistics can be looked at later. So, we force ourselves to focus on what kind of objective was meant for this, and that activated this model," Kumar explains.

Witnessing that data can lead them to make informed decisions, the business aimed to fully leverage its current data to transform magic bricks into a data-driven decision-making company.

He says, As India's real estate market is very unorganized, every information is lying in the pocket of the local agents, it will be a humongous task for a buyer because you would not have any authentic information. Since we are sitting on tons of data, how can we enable our business to make a data-driven decision and help our customer was our aim.

"We restricted to be self-driven. "Someone would send me the data," is the kind of thought process we had. So, an important factor for me was that whichever BI tool I had to select, it had to be user-friendly and not complex. The second factor was to choose what kind of richness that tool could give as a data presentation. With every tool having its pros and cons, we selected features that could be most beneficial. That was vital for us to bring in the data-driven culture," Kumar observes .

Read Also- CIOs in India find data-driven decision-making challengingData model vision With evolving market, Magicbricks is keen on focusing on models that are critical for business. Kumar states, "There are ample opportunities in real estate for data science. But our priority has been to fix those opportunities and then keep on expanding."

Currently, they intends to focus on improving its in-house recommendation engine, which employs a unique algorithm to offer customers the properties most relevant to their selection by combining the power of collaborative and content-based suggestions. The engine contains built-in tweaking that alters the algorithm output regularly. Multiple models make suggestions using various methodologies, and their platforms pool the results based on a dynamic allocation to different models based on the models' day-to-day performance.

"The platform ensures that sufficient recommendations are generated for each user by topping up with the rule-based recommendations in the absence of the required historical activity of the users. We want to invest more in that recommendation engine to get more Click Through Rate (CTR). So, if we have a 15% CTR right now, we want that to go to 30%-80% of CTR. We are working on it," he continued.

In addition, they are doing Proof of Concept (POC) for every technology that is on the market but will delay the adoption. "For example, we evaluated Virtual Reality (VR) three years ago but we haven't done anything on VR because we realize that VR may not be acceptable in the Indian market for now because if you develop it, it will be a fancy word for you but there will be no consumer using it. So, we keep a check on what technology is coming and what technology will work for our business advantage, Kumar reasoned.

Visit link:

Magicbricks aims to strengthen data science focus; 50% team expansion by 2023 - ETCIO