Sumukhi Pandey

Business Intelligence & Analytics professional

Talent

GurgaonMember since 3 June 2026

Power BIDAXTableauPower QueryAdvanced ExcelSQLETLData ModellingData Validation & ReconciliationRequirement GatheringRoot Cause AnalysisPythonPandasNumPyScikit-learnMatplotlibSeabornSnowflakeOracle OCIAWSAzureGenerative AI

Hire this person

Send a job offer directly to this candidate

Power BIDAXTableauPower QueryAdvanced Excel

About

Business Intelligence & Analytics professional with hands-on experience building SQL ETL pipelines, designing Power BI & Tableau dashboards, and automating recurring and ad hoc reports using Python (Pandas, NumPy, Matplotlib, Seaborn). Skilled in complex SQL (joins, CTEs, subqueries, query optimisation), data validation, reconciliation, and quality checks, and data modelling (fact/dimension tables, normalisation). Proficient in DAX, Power Query, Advanced Excel; familiar with Snowflake, Azure, AWS.

Translates complex findings for technical and non-technical stakeholders. Oracle OCI Generative AI Certified · CGPA 9.43 · Dean's Honor List.

Experience

LLM Intern - Ethara AI - Gurugram

(2026-02 - 2026-06)

Executed structured AI model training workflows — annotated, labelled, and curated training datasets across multiple task types; applied prompt engineering techniques to guide model behaviour, improving output quality and alignment with business performance standards across iterative training cycles.
Led quality reviewing of 2,000+ model outputs across 10+ evaluation categories, maintaining ~95% benchmark accuracy; identified systematic failure patterns, documented findings with root cause analysis, and delivered targeted remediations that measurably improved model reliability and output consistency.
Executed structured data validation, reconciliation, and quality checks across 2,000+ AI model outputs (10+ quality categories) maintaining ~95% benchmark accuracy; identified systematic failure patterns and documented analytical methods, logic, and assumptions to ensure consistency and BI governance compliance.
Automated repetitive reporting workflows using Python (Pandas, NumPy); conducted hypothesis-driven root cause analysis across 500+ prompt variants reducing error rate ~25%; presented findings to senior leadership driving consensus on KPI definitions and quality standards.
Gathered and documented reporting requirements from product and engineering stakeholders; translated business needs into scalable analytical frameworks, communicating findings clearly to both technical and non-technical audiences.