I am a senior data scientist at Reed.co.uk, UK's leading recruitment agency. On a daily basis, our team solves problems in recruitment-tech -- related to information retrieval (search/recommendations) and natural language processing. Prior to being at Reed, I was a postdoctoral research fellow at the University of Surrey, where we were focusing on problems related to AI-driven intelligent dental disease detection. I had completed my PhD with Prof. M. Narasimha Murty and a group of researchers and graduate students at the Topic Analysis and Synthesis Lab (TASL) in the Computer Science and Automation (CSA) department, Indian Institute of Science (IISc), Bangalore, India.
My recent research focuses on a few problems in hypergraphs per se (higher-order relation representation and prediction models) and their applications to life sciences. Undirected hypergraph models also serve as pointers to deep set-embedding — another of my areas of research. I have also been interested in explainability in AI in general, and hypergraph/graph neural networks in specific – mostly attention models; apart from that, we are also working on some novel information flow models for deep learning in general.
In the coming future, I would like to explore the frontiers of artificial intelligence, be it a core field (e.g., traditional/modern learning theory, explainability, etc.) or an application oriented direction (e.g., to NLP, computer vision, social networks, etc.), or domain specific (e.g., recruitment-tech, medical, education, etc.). My experience at IISc has made me develop a strong research acumen, which includes pedagogical skills, designing solutions to practical problems, and implementing solutions (I prefer Python, MATLAB, Java, and C/C++ – in that order – but I am a language-agnostic programmer).
I have been a part of TASL since 2010 now, first as a masters student during 2010–2012, and then as a doctoral researcher since 2012. For the past four years, I have been focusing on Machine/Deep Learning solutions to problems in Network Science involving Complex Networks of a Higher Order — Hypergraphs. We have worked on six different projects, ranging from Theoretical Analyses of the Effect of Hypergraphs on Pairwise Links, to Predictive Models for Hyperedges, both Unipartite and Bipartite. Two of our works are in the published state right now, in peer-reviewed proceedings of conferences, and the rest are pre-prints under review, including my PhD thesis. During 2012–15, I was working on several different projects: Markov Logic Networks, Automated Theorem Proving, StackOverflow Edit Prediction, Citation Networks, and causal analysis thereof.
For my masters (2010–12), I had worked on Topic Models for Natural Language Processing – on Text based Music Recommendation to be specific – wherein we had used variational Latent Dirichlet Allocation (LDA) to build an unsupervised mood-based multi-sentiment detector using text lyrics of songs.
On the academic end, I was fortunate to be a part of many interesting courses taught at IISc, of which Linear Algebra and Real & Functional Analysis by Prof. Vittal R. Rao, Probability & Statistics by Prof. Ambedkar Dukkipati and Prof. Indrajit Bhattacharya, and Computational Methods of Optimization by Prof. Chiranjib Bhattacharyya were some of the first courses I had credited, which built a strong academic foundation for machine learning. An unusual course (for me) was on Real Analysis taught by Prof. Gautam Bharali, which questioned my very understanding of Mathematics, and introduced me to the world of rigor and abstraction. The very first core courses on Information Retrieval by Prof. M. Narasimha Murty and Probabilistic Graphical Models by Prof. Indrajit Bhattacharya made me taste the true flavours of the sweet-sounding AI (which were bitter at times, owing to their complexity). The most interesting courses were the ones on Natural Language Understanding and Cognition & Machine Intelligence by Prof. C. E. Veni Madhavan, and Discrete Structures by Prof. Dilip P. Patil — the former ones for their exciting experiments on real text corpora and on an eye-tracker machine, and the latter one for its sheer role in enhancing my mathematical reading, writing, and pedagogical skills.
Apart from learning, I have also worked as a teaching assistant for two courses: one on Discrete Structures by Prof. Dilip P. Patil in 2014, and the other on Linear Algebra by the same professor offered as part of the National Programme on Technology Enhanced Learning (NPTEL) in 2015. Further, I had undertaken several voluntary informal activities in the CSA department: assisting the CSA Web Team for a year, organizing a Mathematics Discussion Club for a year, handling the Opportunities at CSA desk during open-day events, delivering mini-lectures on WordNet and its Applications, Number Theory, Real Analysis, Probability Theory, Mathematical Logic, Linear Algebra, and Matrix Factorization Techniques.
Currently, I am completely involved with the domain of recruitment and the application of artificial intelligence therein. The topic of recruitment and employment is one of those closest to my heart, and I genuinely strive to make contributions in this area. At Reed.co.uk, I'm able to see the impact of my work reach thousands of job-seekers and recruiters, which is something I am proud of.
I've also worked as a Senior Data Scientist at MiQ Digital India Pvt. Ltd. from March 2020 to September 2021, wherein we designed, implemented, and deployed end-to-end machine learning solutions for serving online advertisements intelligently. My experience as a technology trainer spans through 7–8 years across areas such as Algorithms & Data Structures, Stochastic Algorithms, Statistical Learning Theory, Machine/Deep Learning, and its applications to Communication Networks, the Web, Text & Natural Language Processing, and Computer Vision. For around 3 months, I also freelanced for a space-technology start-up as a R&D consultant & developer, wherein I was designing and developing programs to analyze trajectories of satellites in real time.
Some experiences that shaped up my journey as a professional have also been a couple of internships. To start my career in machine learning, I worked for six months (the first half of 2010) as an intern at the Center for Artificial Intelligence and Robotics (CAIR; a Bangalore-based subsidiary of the Defence Research and Development Organization (DRDO), Governement of India), where we worked on Clustering Text Documents using linguistic structures called lexical chains. The next internship I had been a part of was at Wipro Technologies, Bangalore in the year 2015, where for a research project on IT automation, we had designed ML models that would work on the intersection of NLP and Heterogeneous Information Networks.
In the long run, I dream of two things: