Sandeep Reddy - Data Scientist

Overview

Data Engineer (Middle) specializing in production-style data warehouse and ETL pipelines with an emphasis on correctness and validation. The strongest proven skill is building validated ETL and BI-ready warehouse pipelines, as shown by the standardization, fact and mart build scripts and the automated data-quality execution (examples: src/retailiq/warehouse/standardize_remaining.py, src/retailiq/warehouse/build_fact_orders.py, src/retailiq/quality/run_data_quality.py). There is no public evidence of machine learning model development, custom training loops, advanced efficiency engineering or ML lifecycle tooling.

Phone

Timeline

Data Engineer · Middle

Mar Symantech Private Limited · Full-Time

Sep 2023 to Present 3 Years 2 Months In office

Built and maintained ETL/ELT pipelines using Python and SQL to load structured and semi-structured production data into cloud databases and warehouses. Implemented data-cleaning steps such as schema validation and deduplication to protect downstream analytics from inconsistent records. Scheduled and monitored batch jobs with Airflow and cron, diagnosing pipeline failures and resolving production data-processing issues.

Python

SQL

Airflow

Presidency University

Bachelor's Degree · Data Science

2019–2023 Bengaluru, Karnataka

Projects

RetailIQ — Modern Data Warehouse & Analytics Engineering Platform

Designed incremental analytics models for a modern data warehouse with dimensional modeling and slowly changing dimensions (SCD). Produced BI-ready data marts and semantic views using ELT transformations and semantic layer definitions. Delivered a Streamlit dashboard after validating data quality and integration tests.

Cloud-Native Lakehouse, CDC & AI-Ready Data Platform

Implemented cloud-native lakehouse data flow with raw-to-silver/gold layers and CDC ingestion. Built and tested CDC logic using PostgreSQL pgoutput with correct ordering, deduplication, idempotency, and checkpointing. Added extensive automated validation, data-quality checks, and reconciliation tests across multiple datasets/products.

Enterprise AI-Powered Real-Time EV Fleet Data Platform

Processed large volumes of EV telemetry in a real-time pipeline to produce feature windows and AI-ready samples. Used Kafka and Spark for streaming/compute orchestration and stored data in PostgreSQL. Exposed results via a FastAPI service and packaged the solution with Docker and Kubernetes.

Technical skills

Languages

2

Python

SQL

AI/ML

5

Airflow

Machine Learning

Pandas

Streamlit

Scikit-learn

DevOps

3

AWS

Amazon S3

Kubernetes

Other

19

PostgreSQL

SQLAlchemy

DuckDB

FastAPI

Apache Iceberg

Amazon CloudWatch

Spark

Apache Kafka

GitHub

GitHub Actions

Docker

Terraform

Rest API

pySpark

CI/CD

Analytics

ETL/ELT

IAM

Dimensional Modeling

Middle AI/ML Engineer Confidence: Medium Data-centric

Data Engineer (Middle) specializing in production-style data warehouse and ETL pipelines with an emphasis on correctness and validation. The strongest proven skill is building validated ETL and BI-ready warehouse pipelines, as shown by the standardization, fact and mart build scripts and the automated data-quality execution (examples: src/retailiq/warehouse/standardize_remaining.py, src/retailiq/warehouse/build_fact_orders.py, src/retailiq/quality/run_data_quality.py). There is no public evidence of machine learning model development, custom training loops, advanced efficiency engineering or ML lifecycle tooling.

Model Architecture & Training

How well models are designed and trained

Not evidenced in public code

Data Pipeline & Feature Engineering

5/10

How data is prepared for models

Strong, production-oriented data engineering: ETL standardization, dimensional fact and mart construction, and row-level reconciliation are implemented with explicit validation checks.

Evidence

retailiq-modern-data-warehouse/src/retailiq/warehouse/standardize_remaining.py: table creation, standardization INSERTs and row-count reconciliation per-table

retailiq-modern-data-warehouse/src/retailiq/warehouse/build_fact_orders.py: fact population SQL, referential integrity checks and row/grain validation

retailiq-modern-data-warehouse/src/retailiq/marts/build_seller_fulfillment_mart.py: mart build SQL with validation of measures and relationships

Experimentation & Evaluation

3/10

How results are measured and tested

Basic experimentation and evaluation practices focused on deterministic data-quality checks and JSON evidence artifacts; there are programmatic validations and output artifacts consumed by the dashboard but no ML experiment tracking or ablation experiments.

Evidence

retailiq-modern-data-warehouse/src/retailiq/quality/run_data_quality.py: executes reconciliation, grain and quality checks and writes data/output/data_quality/data_quality_execution.json

retailiq-modern-data-warehouse/dashboard/app/main.py: load_evidence function and dashboard consumption of data-quality, governance and observability evidence

MLOps & Deployment

1/10

How models are shipped to production

Minimal deployment or serving engineering for models; there is a Streamlit dashboard for observability and health, but no model serving, versioning or drift monitoring artifacts.

Evidence

retailiq-modern-data-warehouse/dashboard/app/main.py: Streamlit application that connects to PostgreSQL and displays platform health and validation metrics

Computational Efficiency

1/10

How efficiently computing resources are used

Little to no evidence of computational efficiency engineering such as GPU optimization, batching, quantization or profiling; SQL is written to be correct and validated but no measured efficiency work is present.

Evidence

retailiq-modern-data-warehouse/src/retailiq/warehouse/build_fact_orders.py: SQL aggregations and computed delivery duration calculations indicate attention to correctness but not explicit efficiency benchmarking

Research Depth & Innovation

Depth of research and new ideas

Not evidenced in public code

Industries

Commerce· Middle

Data & Analytics· Middle

Technologies

SQL· since 2023 · Middle

Terraform

Spark

Airflow· since 2023

Apache Iceberg

GitHub Actions

Scikit-learn

CI/CD

AWS

Docker

Kubernetes

pySpark

GitHub

Amazon S3

IAM

Amazon CloudWatch

Machine Learning· since 2024

AWS· mentioned only

Cloud· mentioned only

GitHub· mentioned only

Recommendations

Lead development of ETL and lakehouse ingestion pipelines that require strong data-quality and reconciliation controls.
Implement or extend automated data-quality, observability and evidence pipelines that feed BI dashboards and validation artifacts.
Develop BI-ready semantic layers and Streamlit dashboards that surface validation evidence and business KPIs.
Integrate checkpointed incremental/CDC processing and replayable ingestion pipelines with robust idempotency and evidence artifacts

Repositories

The developer's experience in this domain has been verified based on AI analysis of the following repositories:

retailiq-modern-data-warehouse cloud-native-lakehouse-cdc-platform cvsandeep0105-netizen

Middle Data Scientist Confidence: Medium Data Engineer

Data engineer focused on building production-style modern data warehouses and ETL pipelines at a middle level, with a strength in engineering reliable dimensional models and data-quality scaffolding. The strongest proven skill is ETL and data-quality engineering demonstrated by src/retailiq/quality/run_data_quality.py which implements reconciliation, grain checks and produces deterministic execution evidence. There is little public evidence of predictive modeling, MLOps, or automated distributed streaming deployment orchestration in the analyzed human-authored code.

Statistical Rigor

1/10

Correct use of statistics

Minimal statistical rigor; deterministic row-count and simple rule checks are present but no formal assumption testing, uncertainty quantification, or causal analysis.

Evidence

retailiq-modern-data-warehouse/src/retailiq/quality/run_data_quality.py

Data Wrangling & Cleaning

6/10

Preparing and cleaning data

Strong practical data wrangling and cleaning for ETL: trimming, NULL handling, type casts, and defensive standardization are implemented across staging and intermediate layers.

Evidence

retailiq-modern-data-warehouse/src/retailiq/warehouse/standardize_remaining.py

retailiq-modern-data-warehouse/src/retailiq/warehouse/build_fact_orders.py

Exploratory Analysis & Visualization

4/10

Exploring and visualizing data

Exploratory and visualization work is present via a Streamlit dashboard that surfaces KPI charts and tables, but narrative analytics and deep interpretive commentary are light.

Evidence

retailiq-modern-data-warehouse/dashboard/app/main.py

Predictive Modeling

Building models that predict

Not evidenced in public code

Business Insight & Impact

5/10

Turning analysis into business value

Business-focused BI output and KPI design are evident; dashboards and marts link technical outputs to business metrics though explicit cost/error trade-offs or decision-impact analysis are limited.

Evidence

retailiq-modern-data-warehouse/dashboard/app/main.py

retailiq-modern-data-warehouse/src/retailiq/marts/build_seller_fulfillment_mart.py

Reproducibility & Notebook Hygiene

3/10

Clean, repeatable analysis

Some reproducibility practices exist (requirements files, JSON evidence outputs, tests present elsewhere), but there is limited evidence of rigorous environment pinning, data versioning, or automated pipelines.

Evidence

retailiq-modern-data-warehouse/src/retailiq/quality/run_data_quality.py

retailiq-modern-data-warehouse/dashboard/app/main.py

Expertise

Big Data· Middle

Streaming· Middle

Technologies

Analytics

Python· since 2023 · Middle

PostgreSQL

DuckDB

SQLAlchemy

Pandas

Streamlit

ETL/ELT

Recommendations

Develop and extend automated pipeline orchestration (Airflow/Dagster) to run these SQL-based steps as repeatable tasks with CI integration.
Build lightweight data versioning and reproducibility (DVC or manifest-based snapshots) and add pinned environment artifacts to improve portability.
Add richer observability and alerting for quality checks (export metrics, integrate with monitoring) and redact exception outputs to avoid leaking internals in dashboards.
If moving toward ML work, add clear model evaluation notebooks, baseline-first experiments, and documented leakage checks.

Repositories

The developer's experience in this domain has been verified based on AI analysis of the following repositories:

retailiq-modern-data-warehouse cloud-native-lakehouse-cdc-platform cvsandeep0105-netizen

Intern Frontend Developer Confidence: Low Generalist

An intern-level generalist focused on presenting data-engineering work through a static portfolio website. The only observable artifacts are static HTML and CSS pages, but these files are marked as boilerplate/low-ownership and do not provide verifiable evidence of original frontend engineering. There is no human-authored application code, tests, async-state logic, accessibility engineering, or build/tooling evidence to evaluate higher skill levels.

UI Component Architecture

How interface parts are built

Not evidenced in public code

Responsive & Cross-browser

Works on all screens and browsers

Not evidenced in public code

Performance Optimization

Speed of the interface

Not evidenced in public code

Accessibility & Semantics

Usable for everyone

Not evidenced in public code

State Management & Data Flow

Managing data in the app

Not evidenced in public code

UX & Visual Polish

Look and feel quality

Not evidenced in public code

Recommendations

Use the developer for content, documentation and static-site tasks such as updating portfolio pages, copy and visual polish of static HTML/CSS.
Assign them to build small, well-scoped UI tasks under mentorship, for example converting a single static page into a simple component-based React page with accessible markup and tests.
Have them contribute to documentation, engineering reports or story-driven pages that surface data-engineering results while they gain hands-on frontend experience.
Pair them with a senior engineer to implement basic CI checks, linting, and a simple build pipeline for the static site so they can learn build tooling and release workflows.

Repositories

The developer's experience in this domain has been verified based on AI analysis of the following repositories:

data-engineering-portfolio

Job preferences

Work permits: India, United States, Australia, Germany, Canada, United Kingdom
Citizenship: India
Salary: from $30k per year
Employment type: Full-Time
Company size: 501-1000