Conversations: Ex-ISRO scientist on his second innings as a data scientist at Sahaj – INDIAai

From 1984 to 2008, or 24 years to be precise, Ravindra Babu Tallamraju worked on several spacecraft missions at the Indian Space Research Organisation (ISRO). After that, he joined Infosys as a Principal Researcher to work on facial recognition systems. Then, as a Principal Data Scientist at Flipkart, his task was to alleviate issues surrounding address classification, catalogue management and fraud modelling.

Today, Ravindra is the Head of Data Science at Sahaj Software Solutions - a software engineering consultancy company harnessing the potential of data with platform engineering, data engineering and data science.

ISRO provides a number of challenges in deterministic mathematical modelling. As part of Spacecraft Orbital Dynamics activities, one needs to consider various forces that act on the spacecraft and integrate those equations of motion to predict spacecraft position and velocity in future instances.

The position and velocity vectors help derive a host of information such as the spacecraft's visibility over ground stations, ground traces, times of entry of Earth's shadow, etc. Further, camera modelling allows one to compute the coordinates of imaging scans for remote sensing missions.

"While most of the models involve every aspect of mathematics, like forming and solving differential equations, geometrical and numerical modelling, vector algebra, etc., the challenges in probabilistic modelling and machine learning remain limited," said Ravindra.

"This motivated me to look for newer challenges after 24+ years at ISRO," he added.

Sahaj works with its customers on purpose-built solutions involving data and ML. The clientele for AI problems is spread across multiple domains and geographies.

It provides Ravinder with an opportunity to work on a wider scale. The team gets involved in developing/optimising multilingual NLP models, text summarisation, knowledge graphs, natural language queries and contextual conversational engines that enable multiple use cases to be developed leveraging the engine.

In computer vision, Ravinder said, "We work on solutions to generate video summary, sports video analytics, information extraction from videos and images, OCR of noisy scanned documents, and knowledge graphs." Use cases for such AI models include

In mining large datasets, both structured and unstructured, the job is to extract hidden patterns in the dataset. Similarly, when it comes to statistical forecasting, the team has worked on models that forecast using conventional time series, machine learning and deep learning approaches.

All the above-mentioned projects are collaborative in nature, and Ravinder is heading a team of data scientists at Sahaj. As per him, data scientists are inherently researchers. "They love to work on open-ended yet time-bound research problems," he said.

On the one hand, each data scientist has their favourite domain, while on the other side, client projects require time-bound solutions. The need is to find a sweet spot between the two. Ravinder believes that an ideal data scientist would have to work on various problems across multiple data science domains and application areas.

In his own role, he has the responsibility to focus on how to train each data scientist as a well-rounded expert. Further, to inculcate research curiosity, the organisation has a regular Data Science Research forum where the team delivers talks, conference presentations, research publications, and more. Further to this, Ravinder said, "We collaborate with some of the leading academic institutions and undertake joint research projects with expert academicians," he added.

Even though the need for data science is felt across the industry, their problems remain vaguely defined when it comes to skills. Additionally, the team has to look for multiple factors,

At Sahaj, as per Ravinder, any new data scientist is typically paired with experienced data scientists who can sharpen their consulting and problem-solving skills.

Ravinder rightly understands the importance of patents as he himself has a few patents in computer vision. "Associated with having patents, we need an IP team which tracks their possible violations. Sahaj, as an organisation, believes in sharing and open sourcing. As a data science team, we share our knowledge through publications, talks, open source contributions and blogs," said Ravinder.

However, this does not apply to any specific work deliverable or that has been developed as a solution or part of a solution for any of the Sahaj customers.

Moreover, Ravinder sees enormous scope for AI innovations in India. "Over the years, I notice a marked difference in the way academic institutions are open to supporting the private industry through joint collaborations," he said.

In the industry, there are pockets which embrace AI to a great extent, but there are some startups which are still in a very nascent stage. "This requires data science leaders to sensitise on how data science could help provide a business advantage and understand the domain through interactions with the business, product, and engineering leaders to identify the problem landscape," Ravinder concluded.

Read more here:

Conversations: Ex-ISRO scientist on his second innings as a data scientist at Sahaj - INDIAai

Related Posts

Comments are closed.