Looking for a job

Aashish Rajendra Waghela

@aashishrajendrawaghe

Data Scientist at Savani Constructions who cut project budget variance by 7% and prevented Rs. 70 Lakhs in redundant procurement.

India

Message

At Savani Constructions, I engineered a cost forecasting pipeline that reduced project budget variance by 7%. I also created a SKU deduplication engine that prevented Rs. 70 Lakhs in redundant inventory procurement.

At Sigma AI, I built automated evaluation pipelines for enterprise Speech-to-Text models on Google Cloud Platform. My phonetic and dialect discrepancy detection scripts reduced Word Error Rate by 18% across 80+ regional accent categories.

On my Commodity Price Dynamics & Volatility Forecaster project, I designed a time-series forecasting architecture that generated procurement recommendations within 95% confidence intervals. I also hold an MSc in Big Data Science from Queen Mary University of London.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist

Current

Savani Constructions

Mar 2025 - Present (1 year 6 months)

Engineered an end-to-end cost forecasting pipeline using Scikit-learn, reducing project budget variance by 7%. Architected an automated SKU deduplication engine and designed a dynamic resource allocation algorithm improving site operational utilization by 12%.

Python scikit learn Gradient Boosting Random Forest Feature Engineering Cross Validation Fuzzy Matching Levenshtein Distance Google Apps Script Google Sheets Time Series Analysis Resource Allocation Supply Chain predictive modeling

Data Scientist (AI Evaluation Engineer)

Sigma AI

Jan 2023 - Jan 2025 (2 years)

Architected automated evaluation pipelines on Google Cloud Platform to benchmark enterprise Speech-to-Text models, increasing throughput by 17%. Engineered phonetic and dialect discrepancy detection scripts, reducing Word Error Rate by 18% across 80+ regional accent categories.

Python Google Cloud Platform VertexAI Pipelines Cloud Functions BigQuery Google Cloud Storage Dataflow Speech to Text Evaluation Pipelines Data Quality GDPR Compliance Human In The Loop Annotations WER

Machine Learning Intern

Arbuda Infrastructure Projects

Aug 2020 - Aug 2021 (1 year)

Engineered a cloud-native geospatial pipeline on Azure Databricks processing 2M+ spatial coordinates using H3 spatial indexing for flood-risk terrain modeling. Conducted code reviews and implemented ETL optimisation best practices in PySpark/SQL, reducing execution time by 5%.

Azure Databricks PySpark SQL H3 Indexing Geospatial analysis ETL Code Review Optimization

Education

Degrees, certifications, and relevant coursework

Queen Mary University of London

Master of Science, Big Data Science

2021 - 2023

Grade: Distinction

MSc Big Data Science with Distinction. Dissertation on cloud-native anomaly detection using ensemble classifiers, achieving 95% accuracy on 2.9M+ transactions.

University of Mumbai

Bachelor of Engineering, Information Technology

2017 - 2021

Grade: 7.81/10

B.E. in Information Technology with CGPA 7.81/10. Thesis on dimensionality reduction (PCA) and classification models for professional performance benchmarking.