Jan 2024 – Present · 2 yrs 2 mos · Greater Bengaluru Area

• After scaling Docsumo's AI to 1M+ documents/month, I made a strategic decision to advise multiple companies simultaneously as a Fractional AI Leader and Advisor - gaining unique pattern recognition across diverse industries that's impossible in a single full-time role. This dynamic stint has been like getting an executive MBA in AI Strategy and Implementation.
• Currently advising enterprises and startups on AI transformation, coaching technology professionals to upskill in GenAI. Recent clients include Fortune 500 companies (Adobe, 300+ leaders trained), top business schools (Indian School of Business, Master's Union) and Enterprise organizations (IndiaMART).
• 👉 Key Achievements:
• Advised C-suite on GenAI strategy for several B2B & B2C companies, driving AI selection, implementation and adoption
• Led AI implementations including LLM-powered translation and evaluation systems; Agentic AI based recruitment pipelines, RAG-based document intelligence systems, Fine-Tuning LLMs for custom use cases resulting in 40-60% efficiency gains for clients
• Coached 100+ professionals to secure AI/ML roles at Apple, Meta, Amazon, and top tech companies, with 90% job placement rate and $10M+ in salary outcomes
• Created technical content: Published 100+ expert blogs on AI strategy, leadership, and careers reaching 100K+ readers monthly and generating 500K+ impressions; delivered 20+ keynote talks and webinars on GenAI
• Designed cohort-based course training 25+ professionals with 90% securing ML roles in India

Jan 2021 – Jan 2023 · 2 yrs · Bengaluru, Karnataka, India

• As an AI Consultant and Advisor, I partnered with AI startup founders and leadership teams across the US, UK, Europe, and India to accelerate their AI journey. My advisory work focused on:
• 👉 AI Strategy & Roadmapping:
• I advised founders and C-suite executives on defining and executing a cohesive AI strategy, ensuring tight alignment between the technical roadmap and business strategy.
• 👉 Product Development & Innovation:
• Supervised the end-to-end AI product development lifecycle, from initial use-case validation and prototyping to architecting scalable ML systems.
• 👉 Team Building & Scaling:
• Provided expert guidance on building high-performance AI teams, including defining organizational structure and creating hiring frameworks for specialized roles in AI.
• 👉 Thought Leadership & Market Positioning:
• I developed and executed AI content strategies for B2B SaaS startups through expert articles and speaking engagements.

Jan 2023 – Jan 2024 · 1 yr · Bengaluru, Karnataka, India

• Led and mentored an AI organization of 25+ ML engineers to deploy Transformer and LLM-based Document AI models and products, achieving 95%+ accuracy while processing 1M+ documents/month for US-based BFSI clients.
• ​
• 👉 Key Achievements:
• Spearheaded the growth of the AI organization by hiring and mentoring a team of 25+ ML engineers and scientists, managing the complete hiring lifecycle and mentoring
• Established and grew new office in HSR, Bangalore with 30+ employees, creating a culture of cross-functional collaboration and innovation
• Architected LayoutLM, GPT3.5, ChatGPT and open-source LLM based document extraction and intelligence platform, reducing manual processing time by >50%, saving clients >$3M+ annually; improved straight-through processing accuracy to >95%
• Developed synthetic document generation system using template and LLM-based (GPT-4) approaches for training document AI models, reducing data annotation costs by $500K+ annually while improving model generalization by 20% and reducing time to market from >3 weeks to ~3 days
• Led fine-tuning of LLMs (Llama 2, GPT-3.5) for domain-specific document understanding, achieving >85% accuracy on financial documents (15% improvement over base models)
• Established ML platform and MLOps practices including model versioning, A/B testing framework, and monitoring dashboards, reducing model deployment time from weeks to days
• Owned AI organization's technical roadmap and OKR planning aligned with business goals, collaborating across Product, Engineering, HR, Customer Success, and Sales to deliver key initiatives on time
• Drove business growth by leading technical engagement for new client acquisitions, successfully delivering POCs, partnering with Sales, Marketing & Customer success to expand AI solutions for new document types.
• Established and optimized Data Annotation Operations, implementing systems to track and improve key metrics like annotation accuracy and velocity.

Jan 2019 – Jan 2020 · 1 yr · Bengaluru, Karnataka

• Led Conversational AI team for food delivery platform serving 10M+ customers, focusing on customer service automation using Speech Recognition and Natural Language Processing.
• 👉 Key Achievements:
• Spearheaded development of multilingual NLP chatbot (during Covid-19 lockdown) handling 100K+ daily customer queries in English, Hindi, and Hinglish, reducing human agent workload by >20% and saving >$5M+ annually in operational costs
• Architected voice AI system (Hinglish Speech Recognition) for Call Center automation processing 50K+ daily calls with >80% accuracy, reducing average handling time by >25% and improving customer satisfaction scores by >15%
• Led team of 10 Data Scientists delivering high-impact NLP and Speech recognition solutions; mentored 3 junior data scientists promoted to senior roles within 12 months
• Developed and managed AI roadmap for customer service applications, prioritizing use cases by business impact; collaborated with cross-functional VPs in Product, Engineering, Analytics and Operations to ensure alignment with company OKRs
• Drove technical due diligence of 15+ third-party AI vendors, defining and evaluating POCs through data-driven evaluation and ensuring quality standards
• Published Swiggy's first-ever AI research paper at InterSpeech 2020 on code-mixed speech recognition (30+ citations), establishing company's research credibility in the AI community
• Managed hiring and career development for Data Scientists and interns, improving overall team strength, diversity and establishing structured 1-1 mentorship program

Mar 2018 – Sep 2019 · 1 yr 6 mos · Greater Seattle Area

• 👉 Scalable Model Development & Optimization:
• Architected and optimized end-to-end speech recognition models (based on Seq2Seq and Attention), significantly reducing training time on massive speech datasets (~10000 hours) by leveraging distributed multi-host, multi-GPU training on AWS.
• 👉 High-Impact Model Deployment:
• Deployed critical deep learning models for the internationalization of Alexa's NLP capabilities, directly improving the detection of offensive and hate speech in user interactions with Alexa and enhancing customer trust and safety across European markets.
• 👉 Internal Thought Leadership & Influence:
• Founded and authored an influential internal Alexa AI blog to democratize AI knowledge and industry updates. Grew its readership to over 1,500 scientists and engineers across Alexa, AWS, and Amazon orgs, including regular engagement from senior Alexa leadership.
• 👉 Research & Community Contributions:
• Contributed cutting-edge research both internally and externally. Published a paper on sensitive content detection at Amazon's internal Machine Learning Conference and was a contributor to the widely-used 'Dive into Deep Learning' (d2l.ai) textbook.
• 👉 Technical Mentorship:
• Mentored and upskilled software engineers and PhD interns in advanced deep learning, Speech and NLP techniques, fostering the growth of technical talent within the Alexa org.
• 👉 Technology Stack:
• Expertise in deep learning frameworks (PyTorch, TensorFlow, MXNet) and libraries (Fairseq, Tensor2Tensor). Proficient with large-scale distributed training (CUDA) on the AWS cloud platform (EC2) and MLOps tools (Docker). Deep experience with Transformer and Seq2Seq architectures for NMT and ASR.

May 2017 – Jul 2017 · 2 mos · London Area, United Kingdom

• ★ Data Science Fellow at Faculty.ai's highly competitive Fellowship focused on commercial machine learning training for leading academic researchers.

May 2017 – Jul 2017 · 2 mos · London Area, United Kingdom

• ★ Advised the Quantitative Finance and Risk teams to identify high-value clients for risk management at CMC Markets.

Jan 2017 – Jan 2017 · 0 mo · Greater London, England, United Kingdom

• ★ Brand Ambassador (UK) for Revolut, a FinTech startup now valued at ~$33Bn.
• ★ Revolut is a London-based FinTech startup and neobank that offers a secure, mobile-based current account with a pre-paid debit card, currency exchange, cryptocurrency exchange, p2p payments.

May 2016 – Jul 2016 · 2 mos · London Area, United Kingdom

• ★ Mentored a PhD Data Scientist on a project focused on implementing data engineering pipeline for high-density electrophysiological time-series data recorded from the brain.

Jan 2016 – Jan 2018 · 2 yrs · Greater Oxford Area

• ★ Worked as a Sir Henry Wellcome Postdoctoral Fellow in Neuroscience at the Department of Physiology, Anatomy and Genetics, University of Oxford.
• ★ Led an independent research program funded by the Wellcome Trust (GBP 250,000) focused on unsupervised learning and representation of speech sequences in the brain using a combination of electrophysiological, imaging and computational methods.

Jan 2015 – Jan 2019 · 4 yrs · educhroma.com

• ★ EduChroma is a personalized career mentoring startup for students, academics and professionals.
• ★ Profiled in Phenotype, Science magazine of the University of Oxford Biochemical Society.
• ★ Selected as Alpha startup at SURGE, Web Summit's first Tech Conference in India (2016).
• ★ Mentored several students, academics and professionals from leading institutions like University of Oxford, Cambridge, University College London, Imperial College.

Jan 2014 – Jan 2015 · 1 yr · Paris, France

• ★ Investigated unsupervised learning and memory for random noise patterns in the auditory cortex using a combination of in-vivo electrophysiological, computational and anatomical techniques.

Jan 2013 – Jan 2014 · 1 yr

• ★ Rhythm and Time Perception (how does the brain compute time in rhythmic sound sequences)
• ★ Mapped changes in the language network in Stroke patients' brains after therapy
• ★ Working memory (how does the brain encode and remember acoustic features like pitch and time)

Jan 2009 – Jan 2013 · 4 yrs

• ★ Conducted PhD research on how the brain separates meaningful sounds from noise (https://discovery.ucl.ac.uk/id/eprint/1416267/)

Jan 2011 – Jan 2016 · 5 yrs · theassc.org

• ★ As Director of ASSC, I organised several annual ASSC Conferences and supervised the work of various internal committees.

Jan 2008 – Jan 2009 · 1 yr · Greater Oxford Area

• ★ Investigated the pharmacological behaviour of a special type of interneurons in the hippocampus using a combination of whole-cell patch clamp recording and computational techniques.