Hi Sandeep, can you walk us through your journey from being a Software Engineer at Accenture to becoming the Vice President of Machine Learning and Quantitative Analysis at Credit Suisse?
I started my career as a Software Engineer at Accenture. However, I wanted to pursue an advanced degree in AIML and enrolled for a Master’s at the Carnegie Mellon University. In 2011 I joined Credit Suisse as an Analyst within the Investment Bank division post my Masters. I worked on various AIML roles like Client Predictive Analytics, Alternative Data, etc. I was the lead data scientist for Client Predictive Analytics. In 2017 I left Credit Suisse for a role as the Director of Machine Intelligence and Statistical Analysis at the UBS Evidence Lab. I spent the next couple of years designing and developing various AIML models for Geospatial data, Text data, etc. Alternative data was the core focus at Evidence Lab, and this experience shaped me as an expert in leveraging AIML for alternative data. In 2019 I rejoined Credit Suisse as the Head of Machine Learning and Quantitative Analytics, leading a team of ML practitioners and researchers focusing on Innovation within Investment Bank, Alternative data, Sustainability, etc.
I lead the Data Science practice for IB Data and Analytics. My team leverages AIML to generate differentiated content and trade ideas for the Investment Bank. The top 3 challenges have been
1. Process – ML researchers and Data Scientists sometimes think of themselves as Data Artists. Any Data Scientist, including me, would love to get a blank canvas and unlimited time to create the perfect art form (according to the artist). However, from a delivery and impact perspective, this can be quite challenging. Establishing the right process and framework to create the right balance between impact and innovation in this area was something I focused on. Leveraging tools like agile helped us time box some of this without losing creativity.
2. People – Machine Learning and Data Science are some of the hottest skills. Finding and retaining the right talent has been another challenge.
3. Product – Operationalizing AIML models in a very auditable and immutable fashion is a challenge. The absence of enterprise solutions in this space has been a significant challenge. We have leveraged a combination of enterprise products and homegrown custom solutions to ensure our models are explainable and auditable.
Markets are always in Brownian motion. We are always prepared for the unexpected. This will be no different. It will be interesting to see how various sectors recover and the pace of recovery. Data will be a leading indicator, and we are firmly poised to leverage big data and analytics to make data-driven decisions.
There are several. However, two major highlights are the advent of Big Data platforms, making it possible for financial institutions to rapidly mine large amounts of data to make near-real-time decisions. Second, the rise of Alternative Data. Both of these have increased the need or created an opportunity for Software Engineers or specifically AIML specialists in the Finance domain. AIML specialists have started to play a more differentiating role. Today firms are competing to establish themselves as the best Data Science shop. I am lucky to have been a part of both these changes in the industry, the first one at Credit Suisse and the second at the UBS Evidence Lab. This has made my experience unique in the Finance domain.
The financial industry and FinTech are going thru a golden age today when it comes to AIML. All of us are excited to see the disruptions AI can continue to bring in this space. Everything we have seen so far is really just the tip of the iceberg. Other VPNs and Directors in finance that I have been talking to share the same perspective. One lesson that I have learned is that the biggest disruptions to financial products and services may not necessarily come from the Finance industry but rather tech or FinTech. Strong collaborations with FinTech and our peers are the right way to move ahead. This is visible in initiatives like Blockchains or Distributed Ledger Technologies and consortiums like R3.
I think the best advice I got was to focus on what you enjoy doing. One of my mentors and friend, who is now at Google, had given me this advice to be open to new things within your area of interest. This has worked well for me. I was always interested in Mathematics which had led me to Computer Science, AIML, and now Alternative Data. I think it is also important to keep learning and be a “learn it all.” I came across this term in one of Satya Nadella’s interviews. I will end with a learning-related quote from Satya – “Always keep learning. You stop doing useful things if you don’t learn. So the last part to me is the key, especially if you have had some initial success. It becomes even more critical that you have the learning ‘bit’ always switched on.”