Senior Architect Data Science and AI at Tredence Analytics

Architected and delivered the end-to-end KYC Adverse News screening platform, spanning UI ingestion, web search, document scraping, LLM-based summarisation, risk classification, and report generation, ensuring enterprise-grade scalability, security, and reliability. Owned the GenAI and agentic architecture, including LLM model selection, prompt engineering, token and latency optimization, re-ranking, fallback strategies, long-context handling, and asynchronous orchestration with queueing, retries, DLQ handling, and real-time status propagation.

Senior Generative AI Lead at Brillio Technologies

Integrated NokiaGPT (React.js UI) with Jira, enabling real-time query processing via multi-agent orchestration. Built an AI-driven service layer that dynamically routes queries to backend agents (RAG, DB, WriteAPI). Implemented adaptive execution workflows, handling sequential/parallel agent calls based on intent. Enabled real-time streaming responses, delivering incremental results to NokiaGPT UI. Designed and implemented new plugins with defined crawling mechanisms, query-exposed fields, and access controls for project-specific crawlers. Developed indexing logic and scheduled jobs, mapping them to the content database schema for efficient retrieval. Integrated OAuth authentication for JIRA API and established RAG mechanisms with governance and evaluation strategies. Led the technical team of C# developers, data scientists, and QA engineers to execute the project.

Data Science and AI/ML Lead at UiPath

Developed data ingestion and indexing pipelines by creating custom data loaders and node parsers for processing PDFs, technical documentation, and unstructured data. Implemented Llama Index for semantic search, optimizing embedding, indexing, and retrieval for precise information extraction. Designed and integrated custom query engines into an intelligent agent, enabling complex client-specific queries via FastAPI. Designed the architecture for a multilingual machine translation system. Developed ML and NLP models, including decision trees, regression models, SVM, neural networks, and deep learning for various analytical tasks. Created a conversational AI platform for faster invoice approval processing and classified customer service requests for banks. Integrated UiPath products (AI Center, Process Mining, Apps) into a unified Marketplace, enhancing automation capabilities. Established publishing guidelines, curated content, and implemented licensing, improving monetization and user e

Data Scientist at Accenture

Developed predictive models with Pandas, NumPy, and Scikit-learn, optimizing application performance and user engagement. Implemented web scraping and data extraction using BeautifulSoup and Scrapy, improving CAML query accuracy and efficiency. Integrated Salesforce.com with SharePoint, customizing fields, reports, and dashboards, and automating tasks with Python for seamless data flow. Automated workflows and data processing using Python, enhancing List Workflows, Custom Workflows, and Content Iterators for efficient data handling.

Data Engineer at HCL Technologies

Developed a public-facing SharePoint 2013 website, integrating apps with robust backend data handling. Led SharePoint 2010 to 2013 migration, utilizing Apache Airflow, SQL, and NoSQL for seamless data transfer and optimization. Implemented data cleansing, preprocessing, and Power BI dashboards for real-time migration insights and server monitoring.

Data Analyst at Infosys

Provided Level 3 support for critical Microsoft applications, troubleshooting issues using SQL and .NET-based tools. Automated ETL processes and optimized databases, implementing indexing and partitioning strategies for improved performance. Collaborated with end users and higher-level teams, ensuring effective issue resolution, system performance, and database integrity.

Data Science and AI Lead at Brillio Technologies

Developed an enterprise-scale RAG-based AI system integrating Jira, databases, and documents for knowledge discovery, enabling AI-powered decision workflows by combining structured and unstructured data.