Senior ML Technical Product Manager · AI Strategy
At the forefront of LLMOps and enterprise AI — turning complex ML infrastructure into production systems that move business metrics. 7+ years spanning statistical consulting, agentic AI, and ML product leadership at Verizon and Mr. Cooper. 3 US Patents · Chennai, India · Remote-Ready
I'm a Senior AI/ML Technical Product Manager and Applied AI Strategist at Verizon, working at the intersection of LLMOps, data engineering, and product intelligence. My work involves building enterprise-grade AI infrastructure that teams actually use — not demos that die in staging, but production systems that move customer experience metrics and generate IP.
My foundation is statistical — Masters in Statistics from the University of Madras — which means I don't just ship models, I validate them rigorously. I've designed custom statistical test cases that caught model drift before it reached production, and built LLM evaluation pipelines from scratch.
At Verizon, I built ConvoIQ — an LLM-powered conversational intelligence platform analyzing millions of agent-customer interactions. Before that, at Mr. Cooper, I built RAG-powered chatbots, FastAPI PII masking endpoints, and multi-class NLP routing systems. Outside work: 3× US Patent holder (AI-driven customer intelligence, via Verizon).
I offer freelance services in LLMOps Architecture, Statistical Consulting, Agentic AI Engineering, and ML Product Strategy.
Verizon
March 2024 – Present
Built ConvoIQ from scratch — LLM-powered conversational intelligence platform analyzing millions of agent-customer interactions. Led VoC AI product strategy with survey analytics pipelines, NPS enhancement, and churn signal detection. Designed lead management ML suite — propensity scoring, best-time-to-call, channel pairing — augmented with LLMs for business explainability. Authored & secured 3 US Patents covering AI-driven customer intelligence.
Mr. Cooper
August 2020 – March 2024
Built Interaction AI Chatbot using Azure OpenAI & RAG — reduced support query resolution time. Engineered DocAI and Context Splitter for conversation analytics. Developed FastAPI DLP endpoint masking PII at scale. Built multi-class/multi-label NLP classification (DNN) on GCP Vertex AI via Kubeflow. Search embedding via Postgres + ChromaDB + Redis.
Crayon Data
March 2018 – July 2018
Built ML models & scraped itinerary data for a proprietary AI recommendation product. Ran A/B tests for internal product validation. Analysed ad campaign pricing impact across Google & Bing.
Simple Analytics
June 2017 – February 2018
Turned healthcare data into actionable insights using predictive workflows in Alteryx (R & Python). Trained professionals on regression, ARIMA, and Recommendation Systems. Multi-functional startup role — also contributed to hiring and marketing.
01 · Verizon
Built an enterprise-scale LLM-powered platform analyzing 40M+ agent-customer interactions for call quality, sentiment, resolution accuracy, and compliance. Architected multi-layer taxonomy and ontology systems. Owned full prompt garden governance, versioning, and LLM evaluation pipelines. 3 US Patents filed.
02 · Mr. Cooper
Reduced support query resolution time using an intelligent chatbot powered by Retrieval Augmented Generation. Engineered DocAI and Context Splitter modules. Built FastAPI DLP endpoint for PII masking. Search embedding via Postgres + ChromaDB + Redis. Custom sentiment models achieved >90% AUC. Zero PII leakage in production.
03 · Verizon
ML suite predicting customer propensity to convert and determining optimal outbound call times to maximize sales conversion. Designed propensity scoring, best-time-to-call, and channel pairing models — augmented with LLMs for business-readable explainability adopted by sales teams as standard briefing material.
04 · Mr. Cooper
Eliminated manual call routing by building a multi-class, multi-label text classification system automatically routing mortgage customer queries to the right department. DNN trained on call transcripts, deployed on GCP Vertex AI via Kubeflow. Maintained >90% classification accuracy in production.
University of Madras
2018 – 2020
Coursework in statistical inference, multivariate analysis, time series, regression modelling, and applied probability. Foundation for all production model validation and A/B testing frameworks.
University of Madras
2014 – 2017
First Class. Core statistics, probability theory, sampling methods, and computer applications.
Diploma in Computer Applications
2016
Programming fundamentals, database management, and software development basics.
Generative AI Leader
Google AI Essentials & Professional
Microsoft Azure AI Fundamentals
Microsoft Azure Data Fundamentals
Alteryx Designer Core & Professional
Machine Learning using Python
Text Analytics & NLP
Ready to work together?
Book a dedicated 1:1 session — strategy, consulting, or LLM engineering. I confirm within 24 hours.
Open to senior roles, consulting engagements, and interesting problems at the intersection of AI, data engineering, and product strategy. Available remotely or in-person across Chennai.