Atharva UpadhyayData Engineer
Resume , PDF, opens in a new tab
INDEPENDENT THINKING. CONNECTED DATA.PORTFOLIO / 2026

ATHARVA UPADHYAY / DATA ENGINEER

Complex data.
Clear outcomes.

I turn fragmented information into reliable data systems. From property documents to geospatial imagery, I build the pipelines that connect the dots.

Explore my work Download resume

Python & SQL at the core.
Data quality in every detail.

DATA INTEGRATION / SPATIAL ENGINEERING / APPLIED ML

SCROLL TO EXPLORE

PROJECT SCOPEBehind the work,
in numbers.

5,288HTML reports in an extraction workflow
27Standardized commercial-property fields
11,000+Records in a title-shape pipeline
THE CORE STACK
Python PostgreSQL PostGIS Azure OpenCV SQL

01 / SELECTED WORK

A few things
I have built.

Engineering for real-world data.
Explorations in applied intelligence.

Showing 5 projects

NLP & classification03

News Summarization & Sentiment Analysis

An NLP workflow using DistilBERT for sentiment prediction, with automated preprocessing, feature extraction and inference.

  • DistilBERT
  • NLP
  • Python
Project notes

The original portfolio reports 90% sentiment-prediction accuracy for this DistilBERT workflow. This is a previously reported project result, not a new benchmark.

Recommendation systems04

HarmonSync
Music Recommendation

A recommendation pipeline built around feature engineering, similarity-based ranking and an iterative feedback loop.

  • Recommenders
  • Feature Engineering
Project notes

HarmonSync uses similarity-based ranking to recommend music. A feedback loop supports iterative refinement of its recommendations.

Machine learning & finance05

Mutual Fund
Recommendation System

Matching funds to user risk profiles through data-driven clustering and ranking logic.

  • Machine Learning
  • Clustering
Project notes

The original portfolio reports a 25% improvement in match rate through clustering and ranking aligned to user risk profiles. This is a previously reported result, not a new measurement.

Featured case studies describe work at Nimbus Property Systems. Diagrams are illustrative; no source records are shown.

02 / A little about me

The work behind
the useful data.

Part engineering.
Part problem-solving.
Always grounded in the details.

Atharva UpadhyayData Engineer

My work sits where messy real-world information meets reliable data engineering.

I have experience building production ETL/ELT and integration workflows for UK property, planning, title, lease, address and geospatial datasets. Python, advanced PostgreSQL/PostGIS and Azure Blob Storage form the core of my toolkit.

I work through the details that make those systems dependable: conditional matching, deduplication, schema design, spatial validation, referential integrity and production troubleshooting.

THE APPROACH

The details make
the difference.

The work is more than moving data. It is preserving meaning, checking assumptions and making the result dependable.

01 / CONNECT

Meet the data
where it lives.

APIs, HTML, Azure Blob Storage and spatial services, brought into consistent ingestion workflows.

Files · APIs · Cloud
02 / TRANSFORM

Give complexity
a clear structure.

Advanced SQL, schema-aware transformations, conditional matching and relational modeling.

Python · SQL · Models
03 / VERIFY

Quality is part
of the pipeline.

Deduplication, relationship checks, geometry validation and processing logs before downstream use.

Rules · Topology · QA

03 / Experience

Learning by building.

Full résumé , PDF, opens in a new tab

Jan 2025 — Aug 2026

Warwick, UKData engineering

Data Engineer

Nimbus Property Systems Ltd

  • Engineered Python and SQL ETL/ELT pipelines integrating planning applications, Land Registry/title data, UPRNs, leases, addresses, APIs and Azure Blob Storage into PostgreSQL/PostGIS.
  • Built advanced SQL transformations for conditional matching, deduplication and spatial validation. Designed planning-title enrichment that reused validated titles before controlled postcode/UPRN fallbacks.
More about this role
  • Implemented lease/title reconciliation using date-window clustering and rent/floor-area fallbacks to preserve downstream referential consistency.
  • Delivered document-extraction and Scottish title-shape workflows; standardized geospatial ingestion across ArcGIS, WFS/WMS, Cadcorp/GeognoSIS and MapInfo services.
  • Diagnosed type mismatches, invalid geometries, missing relationships and slow SQL paths, using processing logs and QA tables to refine validation rules.
Document extraction Title-shape pipeline

Apr 2024 — Aug 2024

Data integration

Data Analyst Intern

Careervira Education Private Limited

  • Built batch ingestion pipelines from 50+ sources, including Azure Blob Storage, to support scalable data integration and processing.
  • Developed SQL transformations with partitioning, validation and anomaly detection; automated recurring workflows to reduce manual processing time by 40%.

Jan 2024 — Mar 2024

Research & analytics

Research Analyst

Essmart Global

  • Developed predictive models on 100K+ records, using statistical modeling and feature engineering to improve targeting accuracy by 15%.
  • Built data-processing workflows and analytical reports with Python, NumPy and Power BI to support research and decision-making.

04 / The toolkit

Tools with
a purpose.

A practical stack for ingestion, transformation, spatial processing and quality assurance.

01

Python & processing

  • Python 3.12
  • NumPy
  • psycopg2
  • SQLAlchemy
  • OpenCV
  • Regex parsing
  • Automation
02

SQL & spatial data

  • PostgreSQL
  • PostGIS
  • Complex joins & CTEs
  • Query optimization
  • Schema design
  • Relational modeling
03

Azure & integration

  • Azure Storage Accounts
  • Azure Blob Storage & SDK
  • ETL/ELT
  • Batch pipelines
  • API & web-service ingestion
04

Data quality & operations

  • Validation rules
  • Deduplication
  • Referential integrity
  • Topology checks
  • QA tables
  • Root-cause analysis
05

Analytics & reporting

  • Statistical modeling
  • Feature engineering
  • Power BI
  • Analytics-ready datasets

05 / Education

A foundation in
AI & data science.

Nov 2021 — May 2026

Devi Ahilya Vishwavidyalaya

Indore

M.Tech Dual Degree in Artificial Intelligence and Data Science

8.0/10CGPA

A FEW USEFUL DETAILS

Before we
connect.

What kind of work is featured here?

Data integration, commercial-property document extraction and geospatial processing, alongside three original machine-learning projects. The two featured case studies describe professional work at Nimbus Property Systems.

Can I see the source code?

The professional case studies explain the workflows without exposing internal code or customer records. Public work can be explored through the GitHub profile linked below; no unprovided project repositories are linked here.

Visit GitHub , opens in a new tab
Where can I find the full background?

The downloadable resume covers professional experience, technical skills and education in more detail.

Download resume
What is the best way to get in touch?

Email is a direct way to start a conversation about data engineering, analytics or a technical collaboration. LinkedIn and telephone links are also available below.

Email Atharva

06 / Get in touch

A CONVERSATION IS A GOOD PLACE TO START.

Have a data challenge?
Let’s make it useful.

Let’s talk about data engineering, analytics, or turning complex information into something that works.

LinkedIn , opens in a new tabGitHub , opens in a new tab+91 88275 18047

A shortcut to everywhere

Jump to something.

↑ ↓ navigate Enter openEsc close