TRIDEEP DEY DATA ANALYST

GURUGRAM & BANGALORE INDIA

PULL DOWN OR CLICK TO ENTER

PORTFOLIO

I BUILD DATA PIPELINES
THAT TURN MESSY DATA
INTO CLEAR DECISIONS.

Hi, I'm Trideep Dey. I connect fragmented databases, automate manual reports, and build dashboards teams can trust every morning.

PULL DOWN OR CLICK TO ENTER

01 Where I Worked 02 Data Pipeline 03 How I Work 04 Projects 05 About Me 06 Contact

◆ Trideep Dey

Data Analyst · Gurugram & Bangalore

I find the problem in your data,
and build the system to fix it.

I connect messy databases, automate hours of manual spreadsheet work, and build dashboards that leaders can trust every morning.

0% Less manual reporting Replaced spreadsheets with automated feeds

0% Higher client conversion Through clear campaign analytics

0+ Automations shipped Python scripts, ETL jobs, and API syncs

0 Databases unified MySQL + MongoDB into BigQuery

Where I Worked How I Build Pipelines

01 // Experience

Where I Worked

Real problems I solved and the results they created.

Zyla Health · Healthcare · Data Analyst

Current Role

Automating Healthcare Reports & Live Dashboards

The Problem

Patient vitals and clinical records were scattered across MySQL and MongoDB in inconsistent formats. Leadership and doctors had to wait on manual spreadsheet exports to track patient risks and program performance.

What I Did

Built an automated ETL pipeline that pulls from both databases, cleans the data, and loads it into BigQuery every morning. Created live Power BI dashboards that doctors and managers use daily, cutting manual reporting time by 60%.

Python

MySQL

MongoDB

Google BigQuery

Power BI

CredResolve · Fintech · Data Analyst

2023 – 2024

Finding Leaks in Acquisition Campaigns & AI Callers

The Problem

Marketing money was being spent, and AI voice agents were making debt collection calls, but no one had clear tracking on which marketing channels worked or whether the voice bots were converting or failing.

What I Did

Ran the numbers end-to-end and built automated daily performance dashboards. Lowered cost per lead by 15%, increased conversions by 25%, and saved a full working day of manual tracking each week.

Python

SQL

Power BI

Google Cloud

API Data Sync

02 // Architecture

How I Build Pipelines

How raw data flows from different databases into clean boardroom dashboards.

1. Raw Databases

MySQL User profiles & billing

MongoDB Vitals logs & time-series

Incremental Extract

2. Cloud Storage

GCS Bronze Raw Parquet files

Clean & Dedup

3. Warehouse

BigQuery Clean, structured tables

Daily Refresh

4. Reports

Power BI Ready by 9:00 AM daily

Real edge-cases I resolved inside this pipeline:

MongoDB Float Bug

Decimal numbers in MongoDB turned into NULL values during export. Added type-checking guards to keep every ID intact.

Messy Timestamps

Some sources saved Unix seconds, others saved ISO strings. Normalized everything to standard UTC at extraction.

Unexpected Schema Changes

App developers added and renamed database columns without warning. Added schema-drift detection so bad data never reaches reports.

Safe Rollbacks

If a network drop happened halfway through an upload, the pipeline safely rolls back so half-written numbers never pollute dashboards.

03 // Approach

How I Work

A few simple principles that guide everything I build.

Business First, Code Second

I come from commerce (B.Com background). Before writing SQL or Python, I understand the business: what moves revenue, what cuts cost, and what decision someone needs to make. Code is just the tool to get there.

Dashboards Must Answer "What Next?"

A dashboard shouldn't just show what happened yesterday. It should tell the manager or executive what to do tomorrow. If a chart doesn't lead to an action, it shouldn't be there.

Automate Whatever Repeats

Manual reporting wastes time and introduces mistakes. If a report or data pull has to be done more than once, I build a script to do it automatically so the team can focus on real work.

04 // Projects

Personal Projects

Data projects I built on my own to explore interesting questions.

World Bank Data · 216 Countries 21% Power Loss

World Electricity Access & Power Loss

Analyzed global electricity data across 216 countries. Found that Sub-Saharan Africa loses over 21% of its electricity during transmission before it even reaches homes. Built an interactive Power BI map to explore infrastructure gaps.

Python

Power BI

Pandas

View on GitHub →

Cars24 Dataset · Pan-India 40% Price Gap

Used Car Price Trends Across Indian Cities

Analyzed used car sale prices across India. Discovered that the exact same car model with identical mileage sells for up to 40% more in Mumbai compared to Kolkata, purely due to city-level supply and demand differences.

Python

Pandas

Power BI

View on GitHub →

05 // Background

About Me

Who I am and what I work with.

I'm Trideep. I started with a commerce degree (B.Com), which gave me an understanding of finance, costs, and how businesses operate. From there, I got into data and automation—and never looked back.

Over the last 1.5+ years, I've built 50+ pipelines and automations. I spend my time writing clean SQL queries, Python scripts, connecting APIs, and building dashboards in Power BI.

I like taking messy, complicated data problems and making them simple, fast, and automated.

Technologies I use daily:

Python

SQL

Google BigQuery

Google Cloud (GCS)

MySQL

MongoDB

Power BI & DAX