Sakshi
Dhavale
I turn messy transactional and behavioral data into insights — through Power BI dashboards, SQL pipelines, and machine learning models people actually use.
A little context on how I work.
I'm a data and BI professional working across Power BI, SQL, Python, and machine learning — comfortable moving from a raw dataset to a modeled schema to a dashboard a stakeholder can actually act on.
My work spans two data-focused internships and a run of end-to-end projects covering retail analytics, user-engagement reporting, customer segmentation, churn prediction, and computer vision. I care about the full pipeline: getting the data model right, writing SQL that holds up, and building reports and models that hold up under real questions.
The stack behind the dashboards.
BI & Visualization
Database & SQL
Analytics
Machine Learning & CV
Programming
Tools & Platforms
Things I've built end to end.
Mobile Phone Sales — Power BI Dashboard
A retail analytics dashboard tracking revenue trends and brand preference across 15,000+ transactions, with multi-layer drill-through reports for regional performance.
User Engagement & Retention Dashboard
A multi-page Power BI module analyzing behavioral telemetry for 1,000+ users, mapping sleep-logging consistency against subscriber churn indicators.
E-Commerce Sales Performance & Customer Segmentation
A SQL pipeline using CTEs, window functions, and multi-table joins to calculate MoM sales velocity and CLV segmentation, visualized with Python for regional resource-allocation decisions.
Predictive User Churn ML Model
An end-to-end classification pipeline using a Random Forest Classifier to predict subscription churn at 72% accuracy, with feature-importance analysis identifying monthly usage limits as the leading indicator.
LSTM Text Filtering Framework
A deep-learning text-filtering pipeline using LSTM and TensorFlow, improving extraction-classification accuracy by approximately 45%.
Computer Vision Feature Extraction
A Zernike Moments-based feature-extraction pipeline for shape classification, improving classification accuracy by approximately 50%.
PDF OCR Automation
RPA-based text-extraction tooling for scanned transactional PDFs, cutting down manual processing effort.
Where I've worked.
Calm Sleep
Jul 2022 – Sep 2022- Optimized backend SQL queries supporting engagement-tracking modules, improving data retrieval speed by approximately 35%.
- Analyzed behavioral telemetry and subscription usage data to map sleep-logging patterns, surfacing insights for the product team's retention strategy.
- Partnered with product managers to define KPIs for meditation-tracking features and validate new functionality ahead of release.
TecWEG Utility Solutions
Dec 2023 – Mar 2024- Built an automated smart-meter data pipeline using image transfer and OCR, achieving approximately 99% accuracy in automated billing data extraction.
- Designed cloud-based data storage and retrieval architecture for IoT meter-reading data, improving processing efficiency for real-time analytics.
- Collaborated with engineering teams to integrate ESP/GSM-based smart meter hardware with cloud data pipelines.
Background and credentials.
Education
MMCOE
Pimpri Chinchwad Polytechnic
Certifications & Extracurricular
- Python Programming — LinkedIn
- Software Engineer — HackerRank
- Active open-source contributor on GitHub
- Competitive coder on LeetCode & CodeChef
Have a dataset that needs a story?
I'm open to data analytics, BI, and data science roles — always glad to talk about a dashboard, a pipeline, or a model in progress.
Say Hello