Review existing processes, pain points, and inefficiencies in organizations (Technology roadmap, Operator productivity, Downtimes, Scaling issues) Design new processes to minimize human involvement (Data collection, Data architecture, Process Orchestration, UX redesign) Deploy the right LLMs / Small LMs / Multimodal models (Cost-Quality-Perf tradeoffs, Build vs Buy vs Vendor recommendation) Setup touchless model and process feedback loops for continuous learning (LLM fine tuning, MLOps, AI in the loop feedback) Deliver AI business outcomes for client CXOs (Cost and Time reduction, Compliance, Governance, Talent strategy)

Managed Exela‘s product portfolio of B2C SaaS and Enterprise BPaaS offerings for digital mailroom operations that processed more than 2 million physical invoices, checks, claims, and contracts every month. Deployed LLMs (DaVinci, ChatGPT, GPT-4, LLaMA) as production-grade solutions using LLMOps to deliver predictable and repeatable results in Document Classification, Field Extraction, Contract Lifecycle Management, and Customer Support.Built an ecosystem of Data curation guidelines, Annotation frameworks, Model catalogs, and ML Governance to enable citizen development of SOTA AI models that rapidly achieve qualitative (quality of life, transparency, fairness) and quantitative (cost, time, and RoI) business outcomes.

Led the Document Understanding platform team working on SOTA Deep Learning models in Natural Language Understanding and Computer Vision.Managed product and engineering teams on operationalizing this horizontal platform as a [Data Lake -> AI Platform -> MLOps -> Code -> Services -> DevSecOps] pipeline to create AI-enabled solutions in Insurance, Retail, Banking, and Healthcare portfolios.Influenced client CXOs by working with offering and sales leads and demonstrating how our Intelligent Automation solutions drive quantitative (Accuracy, FTE cost, SLA time) and qualitative (no/low-touch, compliance, standardization) benefit levers, thereby improving customer experience for Genpact service lines.Taught on the Genome learning platform as a Master Guru for the Natural Language Understanding and Augmented Intelligence courses, in addition to conducting company-wide training sessions on AI for Business, ML fundamentals, and NLU for Beginners.

Responsible for creating from scratch a variety of NLU products incorporating Domain-specific Sentiment Analysis, Automated Essay Evaluation, Co-reference resolution, Contextual Document Categorization, Concept Identification, Ontology Discovery, and Knowledge Graphs.Led a team of NLP Engineers and Linguists in cycles of planning, implementing, benchmarking, and iteratively improving individual components of an end-to-end Natural Language Understanding pipeline, covering all stages of linguistic analysis – Morphology, Syntax, Semantics, Discourse, and Pragmatics.Managed R&D efforts of the Cognitive Natural Language Understanding platform to stay updated with current trends and ahead of the competition, in addition to closely working with Engineering, Product and Sales leads to quickly and efficiently bring platform capabilities to market.Laid the foundation for a deep linguistics driven NLU pipeline by creating modules with capabilities like Morphological Analysis, Syntax, Dependency and Discourse Parsing, Linguistic Document De-duplication, Semantic Role Labeling, Word Sense Disambiguation, Named Entity Recognition, and Social Media Listening.

Worked in one of the premier Language Technology research groups in the country, and received hands-on training in Computational Linguistics, Natural Language Processing, Machine Learning, and NLP Applications, in addition to research and thesis work on Discourse Analysis, Corpus Linguistics, Dravidian Linguistics, Historical Linguistics, and Sociolinguistics.