I’m Neha Somawanshi, an aspiring Business & Data Analyst who turns messy transaction and business data into clear dashboards and decisions with SQL, Power BI, Tableau, Excel and Python.
I’m Neha Somawanshi, an aspiring Business and Data Analyst who recently completed the Data and Business Analyst training at ExcelR and is now a Business Analyst intern there. I’m a B.Tech graduate in Electronics & Telecommunication (2026, SVERI’s College of Engineering, Pandharpur). An engineering background taught me to look for patterns in noise, and now I apply that to business and transaction data.
My work spans SQL fraud detection on 6.3M transactions and Excel, Power BI and Tableau dashboards. I spot anomalies, validate data with care, and present clear, data-backed findings to stakeholders.
Technical and analytical skills built through projects, internships and certifications.
Documentation and reporting deliverables I can own, from the first requirement to the final dashboard.
Elicit what the business really needs through stakeholder interviews and workshops such as JAD sessions.
Capture the business problem, goals, scope, stakeholders and success metrics in a clear Business Requirements Document.
BRDs written for my Amazon and Zomato projects, available on GitHub.Turn approved business needs into a Functional Requirements Document: features, data fields, business rules and validations.
Break requirements into user stories with acceptance criteria that developers and testers can act on.
Document KPIs, metric definitions, data sources, filters and refresh logic so every dashboard is easy to trust and maintain.
Design clear dashboards with KPI cards, slicers and drill-downs that let stakeholders find answers on their own.
Real datasets and real business questions, plus AI projects from my B.Tech.
Challenge: A naive rule-based flag caught only 16 of 8,213 frauds in 6.3M transactions.
Result: Built a four-rule weighted risk score in MySQL (UDF, views, stored procedure). Fraud recall rose from 0.19% to 94.36%.
View on GitHub ↗Challenge: 9k+ restaurants across 15 countries made it hard to separate real growth from vanity metrics.
Result: Power BI story showing India’s volume vs. quality gap and that delivery boosts reach, not ratings.
View on GitHub ↗Challenge: Category Management needed to find which discounts work and where satisfaction red flags sit.
Result: Analyzed 1,351 products with Power Query, pivots and a Value Score by category, delivered as an interactive dashboard.
View on GitHub ↗Challenge: Raw insurance policy data hid the picture of customers, policy mix, premium growth and retention.
Result: Power BI dashboard on 5,000 policies with KPIs (251.4M total claims, 1.05K avg premium, 33.64% active) and views by status, age, gender, policy type and premium growth.
View on GitHub ↗Challenge: Which app categories to target or avoid, and do pricing, Android version or size affect performance?
Result: Power BI dashboard with Opportunities, Free vs Paid, Android Installation and Size pages, plus an Executive Summary.
More on GitHub ↗Goal: A safe, private space where people can express their emotions and get supportive guidance.
What I built: A Python chatbot using NLP and sentiment analysis to respond supportively and promote well-being.
View on GitHub ↗Goal: Personalized fashion, makeup and accessory advice from real-time face and body analysis.
Highlight: Built with computer vision and machine learning; presented as a paper at ICA6NT 2026.
View on GitHub ↗Completed a three-month insurance-focused business analysis project (Policy Insurance) in ExcelR’s Business Analyst internship programme, applying requirements analysis, documentation and analytical problem solving across the project lifecycle.
AI-Powered Sentiment Intelligence System using Python, Django and Data Science.
Development of an AI agent that tests websites automatically using NLP.
Practical tasks in data analysis and forensic technology.
Certificate of Participation.
AI Integrated Virtual Wardrobe Stylist Using Vision-Based Body Analysis.
Built a risk score that lifted fraud recall from 0.19% to 94.36% on 6.3M transactions.
Power Query cleaning, validation checks and rollback-safe SQL, so insights rest on accurate data.
KPI cards, slicers and drill-downs in Tableau, Power BI and Excel for self-service answers.
Turn findings into clear recommendations for Finance, Category and expansion teams.
Open to Business Analyst and Data Analyst roles as an immediate joiner, and open to relocate to Pune, Hyderabad and Mumbai. Reach out by email or social media and I’ll reply as soon as possible.