I build reliable data pipelines and machine learning models that teams can trust to make decisions.
I started as a systems engineer building ERP software, then spent six years in analytics at
CropBytes, a digital gaming platform, joining as the first analytics hire and later leading
analytics as Senior Data Analyst. Over that time the platform grew from around 2,000 to more
than 650,000 users.
In 2026 I completed an MSc in Computer Science at Aberystwyth University, with a dissertation on
fraud detection. I'm looking for Analytics Engineering, Machine Learning and Python-focused data
roles, particularly in fintech, fraud and financial crime.
650K+users on the platform I led analytics for
60%engagement uplift from a notification engine I built
590Ktransactions in my fraud detection pipeline
Experience
Feb 2021 – Dec 2024
Senior Data Analyst
CropBytes, digital gaming platform, India
Owned analytics for a 650K+ user platform: KPI design, dashboards and data quality across product, engineering and growth.
Delivered a 60% engagement uplift through an in-app notification engine driven by behavioural segmentation.
Built the leadership dashboards used in quarterly OKR planning and roadmap prioritisation.
SQL, Python, BigQuery, Power BI, Redash
Feb 2019 – Jan 2021
Data Analyst
CropBytes, digital gaming platform, India
First analytics hire as the platform grew from around 2,000 to 650,000+ users.
Built behavioural analysis, retention modelling and cohort tracking from scratch.
Defined the company's first KPI framework and reporting standards.
Designed and maintained a custom ERP system covering sales, procurement and employee management.
Led a production payment gateway integration for live transactions.
PHP, MySQL, JavaScript, REST APIs
Projects
Fraud detection pipeline
MSc dissertation, 2026
An end-to-end pipeline on the IEEE-CIS dataset (590,540 card transactions, 3.5% fraud), built to
test three things fraud teams rely on: honest evaluation, feature selection under a tight budget,
and model explanations that stay stable.
dbt and BigQuery transformation layer with a leakage-safe, time-ordered train, validation and test split.
CatBoost models comparing fuzzy-rough (VQRS), mRMR and mutual-information feature selection, scored on PR-AUC with bootstrap confidence intervals.
Measured how stable SHAP explanations are across random seeds.
Re-engineered the VQRS algorithm to cut peak memory from 2.8 GB to 0.7 GB.
Python, dbt, BigQuery, CatBoost, SHAP
Commitly, a study commitment web app
MSc group project, 2026
Product Owner and backend developer in a seven-person Scrum team. I designed the system
architecture, set up the FastAPI and PostgreSQL backend, and built the core commitment and
group logic. The team shipped with 67 automated tests passing and an 8.8/10 user acceptance score.
Python, FastAPI, PostgreSQL, SQLAlchemy, JWT
Public data dashboards
Personal project, 2024 to present
Interactive dashboards of Indian election and budget data. I scrape and structure the data from
official portals and government releases, then publish each dashboard as a single-file web app
with no server.
JavaScript, Chart.js, D3, web scraping
Education
Sep 2025 – Sep 2026
MSc Computer Science
Aberystwyth University, Wales
Dissertation on fraud detection: feature selection, leakage-safe evaluation and explanation
stability on large-scale transaction data. Modules included Advanced Data Analytics, Intelligent
Systems, Modelling, Managing and Securing Data, and Agile Software Development.
CatBoost, XGBoost, scikit-learn, pandas, SHAP, feature selection, model evaluation
Engineering
FastAPI, Airflow, Docker, Git, GitHub Actions, Linux
Analytics and product
KPI design, A/B testing, cohort and funnel analysis, segmentation, Power BI, product ownership
Spoken languages
English (IELTS 8.0), Malayalam, Hindi
Contact
I'm available for full-time roles across the UK and happy to relocate. The best way to reach me
is a message on LinkedIn or an email. My CV and references are available on request.
