Anshika sharma

Senior Data & AI Associate

Mohali 2+ yrs exp 84 · Excellent

About

Senior Data & AI Associate with 2+ years of experience in insurance technology, with hands-on experience in SQL, Python, data analysis, ETL, data validation, reconciliation, anomaly detection, and Power BI. Experienced in analyzing 5,000+ daily transactional records and automating data-driven operational workflows. Hands-on exposure to Generative AI, LLMs, RAG, prompt engineering, LLM APIs, and AI-assisted automation, including customer-facing GenAI chatbot and knowledge-retrieval use cases.

Skills & Expertise (32)

SQL Advanced

8.5/10

Years Exp

Pandas Advanced

8.0/10

Years Exp

Prompt Engineering Advanced

8.0/10

Years Exp

Generative AI Advanced

8.0/10

Years Exp

Python Advanced

8.0/10

Years Exp

NumPy Advanced

7.5/10

Years Exp

Data Cleaning Advanced

7.5/10

Years Exp

Data Transformation Advanced

7.5/10

Years Exp

ETL Intermediate

7.0/10

Years Exp

Git Intermediate

7.0/10

Years Exp

Data Pipeline Intermediate

7.0/10

Years Exp

AWS Intermediate

7.0/10

Years Exp

Power BI Intermediate

6.5/10

Years Exp

Data Workflows VS Code Jupyter Notebook Azure DevOps Query Optimization AI output evaluation llm apis Data Validation RAG Data Reconciliation LLMs CSV JSON REST APIs Advanced Excel ELT Data Integration Data Normalization Data Quality Monitoring

Work Experience

Senior Software Engineer

Yovant Labs

Aug 2024 - Present

Analyzed 5,000+ daily transactional and operational records using SQL and Python to identify trends, anomalies, process gaps, reconciliation issues, and revenue-impacting exceptions, supporting data-driven operational decisions. Developed complex SQL queries using joins, CTEs, subqueries, window functions, aggregations, and conditional logic to extract, transform, reconcile, and analyze data for financial, operational, KPI, and business reporting. Built a Python/Pandas anomaly-detection system that flags about 7–8 data anomalies weekly across daily transactional reports, autonomously resolving around 60% without manual review, reducing manual monitoring effort by 60+ hours and shifting team focus from reactive checks to root-cause analysis. Designed an LLM/RAG investigation layer that surfaces historical resolutions and root-cause context for flagged anomalies in seconds, cutting investigation time on recurring payment issues and reducing repeated root-cause analysis across the team. Supported evaluation and integration of pretrained LLMs for enterprise analytics and automation through prompt engineering, structured data preparation, RAG-based knowledge retrieval, response validation, and AI output evaluation, with secure handling of insurance data. Contributed to a customer-facing GenAI FAQ chatbot (RAG + prompt engineering, integrated into the Duck Creek UI) that answers common insurance queries directly, reducing routine support inquiries reaching human agents.