Aug 2021 – Jan 2022 · 5 mos · Guildford, England, United Kingdom · On-site

• As a Research Fellow at the University of Surrey, I led a 6-month project in collaboration with the Royal Surrey County Hospital under the Higher Education Innovation Fund initiative "Automated Dental Disease Detection with Trustworthy Deep Learning". My work focused on developing advanced object detection methods -- including mathematical extensions to the YOLO framework -- for identifying dental diseases in X-rays. I also designed and deployed a VGG-like annotation tool to support data labelling and model development.
• I was responsible for the full data lifecycle, from acquisition and secure storage to processing, and authored a detailed Data Protection Impact Assessment (DPIA) to ensure compliance with NHS and hospital governance standards. I further contributed to Bayesian crowdsourcing approaches for handling imperfect labels and delivered a prototype demonstrating system feasibility.
• In addition to research delivery, I mentored PhD students in applied ML workflows and experiment design, and coordinated with academic and clinical partners to align research outputs with healthcare needs. This fellowship served as a post-COVID research sabbatical, reinforcing my expertise in computer vision and trustworthy AI before returning to industry leadership.

Sep 2020 – Aug 2021 · 11 mos · Bengaluru, Karnataka, India · Hybrid

• As the Founding Data Scientist at Factors.AI, a marketing analytics platform comparable to Google Analytics, I built the company's data science capability from scratch. Working in a GoLang-based environment without access to Python's ML ecosystem, I wrote custom large-scale data mining and AI code to deliver core product features.
• Key contributions:
• Developed multi-touch attribution pipelines (Shapley, Markov) to quantify campaign ROI.
• Built conversion and retention prediction models to guide acquisition strategies.
• Implemented multi-dimensional histogram and hyperloglog-based methods for scalable trend analysis.
• Created a natural language query bot enabling marketers to directly interrogate the data.
• Produced week-over-week change analyses and interventional insights for campaign optimisation.
• Contributed to research outputs, including patents and publications.
• This role gave me end-to-end ownership of the analytics and AI layer powering the product, from PoC to production, while collaborating closely with the founders on roadmap and strategy.

Jul 2020 – Sep 2020 · 2 mos · Bengaluru, Karnataka, India

Mar 2018 – Sep 2019 · 1 yr 6 mos · Bengaluru, Karnataka, India

• At MiQ Digital, I made my first formal transition from academia into industry, developing both a product mindset and the ability to translate advanced ML into business outcomes. As Senior Data Scientist, I built and productionised ML solutions for ad targeting and optimisation, while learning how profit, R&D, and product strategy intersect in practice.
• Key contributions and learnings:
• Built embedding pipelines ("ad2vec", "user2vec", "publisher2vec") for user, ad, and publisher modelling, driving improved targeting and forecasting.
• Designed multi-armed bandit optimisation models for dynamic ad budget allocation.
• Developed fraud detection systems to improve campaign integrity and client trust.
• Drove end-to-end pipelines for ingestion, transformation, and alerting of high-volume ad data.
• Developed strong skills in stakeholder management, user empathy, and pitching technical ideas in layman terms to solve client-facing problems.
• Mastered the art of prioritisation, connecting profit with R&D investment and aligning solutions with product and commercial needs.
• Became a master translator and documenter, bridging researchers, engineers, and business teams.
• This role was a defining break from my academic journey, teaching me how to combine cutting-edge ML with industry pragmatism, and shaping my ability to operate at the interface of AI, product, and business value.

Jan 2017 – Apr 2017 · 3 mos · Indian Institute of Science, Bangalore

• Course: Linear Algebra
• Instructor: Dr. Dilip P. Patil
• Link: http://nptel.ac.in/noc/individual_course.php?id=noc17-ma04

Feb 2016 – Feb 2017 · 1 yr · Bangalore

• Worked under Dr. Anurag Srivastava (then Senior VP, Wipro) on formulating a model to automate IT Service Management systems. We designed the architecture of a model that would handle raised tickets, and suggest/implement standard operating procedures, which involved heterogeneous information networks.

May 2013 – Sep 2020 · 7 yrs 4 mos · Bengaluru, Karnataka, India

• As a trainer and training material developer,
• Handled end-to-end corporate training pipelines involving deep/machine learning (D/ML)
• Trained 40–60 sized teams on: D/ML in Software; DL for Web; ML for Communication Networks; Mathematics in ML; Statistical Data Analysis

Aug 2012 – Dec 2020 · 8 yrs 4 mos

• As a full-time Research Scientist at IISc, I advanced the theory and applications of machine learning, natural language processing, and hypergraph models, under the supervision of Prof. M. Narasimha Murty. My doctoral research focused on hypergraph networks -- their structure, analysis, and predictive models -- resulting in multiple peer-reviewed publications and collaborations.
• Key contributions and experiences:
• Hypergraph Research: Designed and analysed learning algorithms for hypergraph models, extending graph ML into higher-order structures.
• Mathematical Foundation: Applied advanced probability, statistics, and measure theory to develop rigorous machine learning methods.
• Publications & Collaborations: Published in international venues and worked with global research groups.
• Teaching & Mentoring: Served as teaching assistant for advanced ML/AI courses; delivered tutorials and lectures; mentored Master's students annually through their theses; and guided junior researchers.
• Community Leadership: Organised student-led initiatives, technical wikis, summer schools, and departmental outreach activities.
• Breadth of Training: Completed advanced coursework across mathematics (algebra, probability, real analysis, logic) and advanced ML (information retrieval, NLP).
• This role shaped me as an independent scientist and mentor, combining abstract mathematical reasoning with applied machine learning, and instilled values of team-building, intellectual honesty, and devotion to truth-seeking that I continue to carry into my leadership roles today.

Aug 2010 – Jul 2012 · 1 yr 11 mos

• My Master's at the Indian Institute of Science (IISc) was the most formative stage of my journey as a researcher and AI scientist. I undertook three times the standard coursework load, covering intensive mathematics (linear algebra, probability, real analysis, optimisation) and the breadth of artificial intelligence, machine learning, deep learning, information retrieval, and natural language processing. This immersion built the foundational depth that I continue to draw upon throughout my career.
• My thesis research, "Sentiment-driven Topic Analysis of Song Lyrics", combined probabilistic topic modelling (Variational LDA) with sentiment lexicons (SentiWordNet) to explore sentiment-aware content analysis at scale. In the process, I curated and built large text corpora, implemented statistical learning pipelines, and validated results experimentally. This work led to my first peer-reviewed publication and gave me early exposure to bridging mathematics, data, and human-centered insights.
• Equally important was the culture of research and teaching at IISc. I actively engaged in seminars, collaborative projects, and peer-learning groups. I built habits of discipline, intellectual honesty, and persistence that shaped my scientific outlook. The intensity of coursework and research also prepared me to transition naturally into doctoral research, while giving me the confidence to later mentor students of my own.
• This role was far more than an academic degree -- it was a full-time research experience that gave me the mathematical maturity, AI breadth, and research skills that underpin both my doctoral contributions and my later industry leadership in AI.

Mar 2012 – Jun 2012 · 3 mos · Bengaluru, Karnataka, India

Jan 2010 – Jun 2010 · 5 mos · Bangalore

• Worked under Dr. Dipti Deodhare (then Scientist ‘F’, Centre for Artificial Intelligence and Robotics, Defence Research and Development Organization) on using a semantic representation, viz., “Lexical Chains”, for documents, detecting topics using word-word semantic similarity, and finally clustering them using a graph clustering technique.
• Resulted in a publication in International Conference on Natural Language Processing 2010, titled:
• "Semantically Driven Soft-clustering of Documents using Lexical Chains"