Sarthak Mistry | Data Engineer

About Me

I'm a data engineer focused on Batch & Real-Time Pipelines, Cloud Data Infrastructure, and Data Modeling. I help teams turn raw, messy data into reliable, well-governed systems that power real decisions.

With a strong foundation in production-grade tooling and a detail-oriented approach, I build data platforms that are not just functional but scalable, observable, and maintainable — from ingestion to insight.

View Resume

Your Photo

01 Experience

Jan 2026 – Present

Graduate Research Assistant

Indiana University, Bloomington

Designing a unified API layer for OpenBT to enable load/save operations across cloud storage services (Arweave).
Conducting validation experiments to benchmark code reliability and accuracy for the C++-to-Rust translation of OpenBT.

C++RustArweave

Aug 2025 – Present

Graduate Teaching Assistant

Indiana University, Bloomington

Mentored 100+ students in Management Access and Use of Big Data (INFO-I 535)
Designed 10+ assignments covering GCP, cloud computing, virtualization, data pipelines and modeling

Big DataGCPCloud

Jan 2025 – Sep 2025

Data Engineer

Project 990

Automated Pipeline for Ingesting yearly county level data from Bureau of Economic Analysis, Bureau of Labor Statistics and American Census Survey.

Python

Jul 2022 – Jul 2024

Data Engineer 1

Piramal Capital & Housing Finance, Mumbai

Developed batch pipelines ingesting data from Salesforce, MongoDB, SQL Server into Snowflake via Spark & Airflow
Improved regulatory reporting efficiency by ~40% automating a 150-feature Snowflake data mart
Reduced PII exposure risk with automated masking, 93% precision across 1M+ records
Built real-time CDC pipeline using Debezium, Kafka, and Spark Streaming
Automated economic data scraping using Python and Selenium, improving operational efficiency by 73%.
Increased ingestion speed by 46% automating AWS Lambda S3-to-Snowflake loads

SnowflakeSparkKafkaAirflowAWSDebezium

Oct 2021 – Mar 2022

Intern — Dex Analytics

NSEIT, Mumbai

Accelerated delivery cycles by engineering a CI/CD pipeline for dbt models in PostgreSQL, automating pre-aggregated data modeling and reducing dashboard query latency by 40%.
Reduced data ingestion errors by 25% by implementing automated dbt schema validations and data quality tests for candidate demographic datasets prior to Kibana visualization.

dbtPostgreSQLKibanaCI/CD

02 Projects

Architecture Diagram

E-Commerce Medallion Pipeline

GitHub

Airflow · Astronomer · DBT · Snowflake · Great Expectations

End to end data pipeline on Snowflake with a 4-layer medallion architecture (RAW → Staging → Data Vault 2.0 → Star Schema) across 17 dbt models.

50+ data quality expectations across 7 suites using Great Expectations
Data Vault 2.0 layer with 3 hubs, 2 satellites, and 1 link table using MD5 hash-diff change detection and incremental loading.
2 Airflow DAGs via Astro CLI and Cosmos, rendering each dbt model as an individual task.
Data Mart with RFM customer segmentation, product performance tiers, MoM growth, rolling revenue windows, and running aggregates.