k

kaviya s

About

Detail

Tamil Nadu, India

Timeline

work
Job
school
Education

Résumé

Jobs verified_user 0% verified
  • Xerago
    Senior Associate AI Engineer
    Xerago
    Jan 2023 - Current (3 years 6 months)
    Design and deploy AI solutions using on-prem LLMs, RAG pipelines, and vector databases.Build scalable ETL and data pipelines for multi-source ingestion, transformation, and analytics workflows.Develop and maintain backend AI services using Python, FastAPI, and MySQL for production integration.Create intelligent reasoning layers for intent detection, NL2SQL, task routing, and tool automation.Collaborate with cross-functional teams to convert business requirements into reliable AI-driven features.Enhance LLM/RAG pipelines through prompt tuning, testing, and continuous accuracy improvements.Develop and update backend APIs to integrate AI functionalities into live products.Monitor, validate, and fix ETL workflows to ensure clean and consistent
  • C
    Automated Data Analyst Intern
    CLEAR QUOTE TECHNOLOGIES
    Jul 2022 - Oct 2022 (4 months)
    Annotated and prepared large automotive image datasets using VGG Image Annotator to support ML model training.Cleaned, validated, and organized data to improve dataset quality and ensure accurate vehicle damage classification.
Education verified_user 0% verified
  • Bharathiar University
    Master of Science in Data Analytics
    Bharathiar University
    Oct 2020 - Jun 2022 (1 year 9 months)
    Honors: 83%
  • Bishop Heber College
    Bachelor of Science in Actuarial Science
    Bishop Heber College
    Jun 2017 - Apr 2020 (2 years 11 months)
    Honors: 74%
Projects (professional or personal) verified_user 0% verified
  • B
    Bharat Matrimony - Matrimony AI Chatbot
    Built a production-grade conversational AI using LLMs + retrieval flows.Extracted knowledge from PDFs, Excel sheets, and website content using custom ETL pipelines.Implemented intent classification, workflow routing, and dynamic context management.Integrated ML scoring models for match suggestions.Improved user response accuracy with embedding-based FAQ retrieval.
  • V
    Virtual Assistant - Matrimony AI
    Built a hybrid LLM + rule-based workflow using RAG and agentic decision layers.Implemented embedding-based task matching using Qdrant with full conversation context.Automated profile insights, compatibility scoring, and event-based task triggers.Integrated backend tools via FastAPI with a RASP-like tool execution pipeline.Designed 300+ user scenarios and created a complete Work Breakdown Structure (WBS) for the VA.
  • J
    Journey Assist - AI Assistant
    Engineered advanced NLP pipelines using on-prem LLMs (Ollama) for automated To-Do extraction, meeting detection, intent classification, and journey summarization.Built a robust LLM reasoning layer using LangChain to deliver context-aware insights from website analytics and email data.Developed ML models for user segmentation, drop-off prediction, and journey-based recommendations improving prediction accuracy across customer journeys.Designed scalable data engineering pipelines to unify Umami, Google Analytics, GSC, IMAP emails, and custom tagging into ML-ready datasets.Delivered a secure, domain-agnostic AI system that reduced manual analysis time by 70% through LLM + ML automation.
  • C
    Conversigo - AI Analytics Chatbot
    Designed and developed end-to-end AI pipelines integrating LLMs, RAG, and Qdrant vector DB for a real-time analytics chatbot system.Built and optimized GA4 → BigQuery → MySQL ingestion and transformation pipelines for scalable analytics retrieval.Implemented NL2SQL workflows, prompt engineering, and SQL validation for high-accuracy data responses.Developed automated analytics report generators with visualizations, anomaly detection, and auto-insights using Python.Improved system performance through query optimization, caching, and model fine-tuning, reducing user query response time significantly.