Headline

Lead(Manager) Agentic AI Developer | M.S. Data Science | LLM | Generative AI | Agentic AI | Graph RAG | Prompt Engineering | RAG | Knowledge Graphs | SQL | AWS, GCP | Databricks | Leadership | Driving Million $ Impact

About

Led the design and delivery of a scalable supervised Agentic AI solution to transform customer complaint handling across multiple business units. The system intelligently analyzes inbound emails—including text, document attachments, and images extracts key attributes, summarizes case history, validates customer information, and auto-raises ServiceNow incidents. Also implemented contextual response drafting capabilities, enabling personalized and efficient customer engagement. Orchestrated the end-to-end architecture using AWS services (Lambda, SQS, Step Functions, Bedrock) alongside GitLab and Atlassian toolchains—ensuring high performance, auditability, and extensibility. Used Anthropic Suite of Models - Claude, Sonnet and Haiku along with Mistral models for GenAI development Managed and scaled a cross-functional team of 7, driving development velocity and AI-first delivery practices in a regulated enterprise environment. Projected business impact: >$2Mn in cost savings over 2 years through automation, improved accuracy. Used Anthropic Suite of Models - Claude, Sonnet and Haiku for GenAI development- 2. Machine Learning & NLP-Driven Revenue Acceleration Directed an enterprise-scale AI solution for B2B lead scoring and product alignment for a UK-based telecom major applicable to banking's commercial/SME segments. Collaborated closely with business stakeholders to co-create use cases and wireframes aligned with strategic goals. Engineered a robust data ingestion and transformation pipeline integrating structured and unstructured data from web scraping, APIs, JSON dumps, and BigQuery. Migrated data through GCS to BigQuery for large-scale processing. Applied advanced Transformer-based NLP models (e.g., BERT variants) and ensemble ML methods to power lead prioritization and contract alignment models. Deployed on GCP using BigQuery, Vertex AI, and Cloud Composer to manage millions of records, delivering production-grade solutions at scale. Business impact: Contributed to a $10Mn+ uplift in EBITDA by optimizing lead targeting, enhancing conversion rates, and streamlining sales operations.

Location

Pune Division

AI stack

cursor

opencode

claudecode

Work experience

Lead(Manager) Agentic AI Developer at Mastercard, Jul 2026 - Present · 2 mos, Pune City, Maharashtra, India

AI Senior Software Engineer - II (Forward Deployed Engineer) at First American (India), Jan 2026 - Jul 2026 · 7 mos, Bengaluru, Karnataka, India, (Agentic AI & RAG Applications, Python, LangGraph, Azure Databricks, Knowledge Graphs) - Design and deploy production-grade agentic AI systems using LangGraph, Knowledge Graphs and RAG, enabling multi-step, stateful reasoning workflows. - Build and iterate on Retrieval-Augmented Generation (RAG) pipelines, optimizing retrieval quality, grounding, and latency. - Evaluate, fine-tune, and serve LLMs and embedding models at scale on Azure Databricks, focusing on robustness, observability, and cost efficiency. - Develop end-to-end AI systems spanning data ingestion, orchestration, inference, and monitoring in Python. - Mentor engineers on LLM systems, evaluation techniques, and responsible deployment.

Data Science Associate Manager at Accenture Strategy & Consulting, Dec 2023 - Jan 2026 · 2 yrs 2 mos, Pune District, Maharashtra, India, • Spearheaded the development of an Agentic AI system automating service incident creation and customer response across enterprise units. Integrated email, image, and document processing with real-time decision intelligence. • Designed and deployed(productionized) scalable GenAI pipelines using AWS (Lambda, SQS, Step Functions, Bedrock, SageMaker, IaaC) and Anthropic Models(Sonnet, Haiku, Mistral) ensuring low-latency. • Mentored a team of 7 data scientists, data engineers, testers, etc; promoted AI-first delivery culture across agile squads. • Resource and time savings led total savings and revenue upliftment of over $2Mn • Engineered and designed an event-driven pipeline for dynamic call summarization and form filling using Anthropic models and prompt engineering in a contact center platform, achieving $1.5M+ operational savings. • Built LLMOps components for prompt versioning, agent monitoring, and feedback integration using AWS S3, Bedrock, Lambda and ServiceNow(for capturing user feedback). GenAI, Applied Research and RFP: • Leveraged LLMs (Claude, Haiku, Sonnet) and foundation models for summarization, validation, and personalization tasks. • Contributed to internal research, POCs, and innovation initiatives using GenAI , RAG and Agentic AI building RFP, BD’s and client proposals using LangChain, LangGraph, Streamlit, etc. • Maintained awareness of emerging research in generative modeling, model context protocol (MCP), Agentic AI techniques and tools.

Data Science Specialist (Consultant/Team Lead) at Accenture Strategy & Consulting, Aug 2021 - Dec 2023 · 2 yrs 5 mos, Pune Division, Maharashtra, India, GenAI Studying and experimenting with GenAI capabilities and various LLM Models Designing prototypes, wireframes to be used for client pitches involving Gen AI Intelligent Revenue Growth (Propensity Modeling) o The problem statement involved identifying and prioritizing B2B leads along with the right set of products to aid in business growth for a large Telecommunications organization in UK. o As part of data strategy, worked with client to understand the problem statement and to build and brainstorm on the use cases as well as design wireframes for better understanding. o Data gathering involved using various structured and unstructured data from multiple sources like News, Company websites, 3rd Party, SalesForce data, etc. which was obtained from web scraping, BigQuery tables, json data loads, API’s, etc. The project involved using transformer-based NLP algorithms along with ensemble ML Algorithms on huge datasets (millions of records) o Designed wireframes to identify lost opportunities and upcoming renewals and further developed BigQuery codes and Qlik dashboards to build the view for sales team, to target specific accounts as per the renewal time frame. o GCP (BigQuery, VertexAI, Cloud Composer, AI Platform) was used to design, create, and deploy the model.

Software Engineer at Avaya, Jul 2019 - Aug 2021 · 2 yrs 2 mos, Pune Area, India, • Video verification in video call – Created a tool which would be used to check video quality, stability in a P2P video call and video conferencing using Computer Vision. This saved cost and time by reducing 1 day of manual work to few minutes. • Smart automation – Execution status was uploaded to a SQL DB. A supervised ML algorithm(based on Decision Trees) was developed to prioritize the testcases for next run. This helped in isolating of issues and faster bug fixes by 30%. • Meeting Summary Generator - A silent participant joined an ongoing meeting and the stream was sent to Google STT to convert the Speech to text . The text was then summarized using various NLP algorithms like TexRank, LexRank and others along with finding action items and provide a meeting summary to the participants. • Insights Web App - Google Analytics and JIRA data of IX workplace containing most used features, issues were extracted, and areas / features of most importance were identified using NER. Tableau was used to visualize the data and recommend areas where most bugs were reported. This resulted in 10% reduction in customer found defects.

Senior Tech Associate at Avaya, Jul 2018 - Mar 2019 · 9 mos, Pune, Maharashtra, India

Management Associate at Avaya, Jul 2017 - Mar 2019 · 1 yr 9 mos, Pune, Maharashtra, India

Education

Master of Science - MS, Data Science, Liverpool John Moores University

Post Graduate Diploma, Data Science, International Institute of Information Technology Bangalore