DATA SCIENTIST & ANALYST

Sri HarshaPeetha.

DATA SCIENTIST DATA ANALYST BI ANALYST

MSc Data Science & Analytics with Distinction — University of Hertfordshire. I build end-to-end ML pipelines, deploy to Azure, and translate complex data into clear business decisions.

sri_harsha_peetha.py
python profile.py
 
# Initialising profile...
 
{
  "name": "Sri Harsha Peetha",
  "location": "Hatfield, UK",
  "degree": "MSc Data Science",
  "grade": "Distinction",
  "stack": ["Python", "SQL", "R",
           "XGBoost", "Power BI", "Azure"],
  "right_to_work_uk": True,
  "open_to_work": True
}
 
Scroll
Projects

Selected Work

Customer Churn Prediction

END-TO-END ML · AZURE DEPLOYMENT · LIVE API

Full MLOps pipeline on the Kaggle Telco Churn dataset. Trained and benchmarked Logistic Regression, Random Forest, and XGBoost — selecting Logistic Regression for deployment based on best accuracy and F1. Tracked all experiments with MLflow and deployed a RESTful Flask API to Azure App Service (Linux, Python 3.10) using Gunicorn. Resolved Kudu environment dependency issues via a custom startup command.

PythonLogistic RegressionRandom Forest XGBoostScikit-learnMLflow FlaskAzure App ServiceGunicornGit
View on GitHub
Logistic Regression — deployed Acc 76% · F1 0.77
Random Forest Acc 75% · Recall 82%
XGBoost Acc 72% · F1 0.73
MLflow Tracking ✓ Active
Azure Deployment X Inactive

Flood Prediction — Ensemble ML

MSC DISSERTATION

XGBoost achieved R² 95.51% and RMSE 0.0106 after hyperparameter tuning — outperforming LightGBM (R² 94.98%) and Random Forest (R² 78.51%). Feature importance analysis identified top environmental predictors, delivering actionable insights for disaster management stakeholders.

PythonXGBoostLightGBMRandom ForestScikit-learnGoogle ColabAgile
XGBoost R² 95.51%
LightGBM R² 94.98%
Random FOrest R² 78.51%
View on GitHub

Superstore Sales BI Dashboard

POWER BI · DAX

Interactive Power BI dashboard on the Kaggle Superstore Sales dataset. Built DAX measures, drill-through filters, and dynamic slicers for self-service exploration by region, product category, and time period — replicating a production BI reporting environment.

Power BIDAXPower QueryData ModellingKPI Reporting
View on GitHub

Heart Disease Risk Analysis

HYPOTHESIS TESTING · R · GROUP PROJECT

Investigated the statistical association between EKG results and heart disease prevalence using chi-square proportion hypothesis testing in RStudio on a clinical dataset (n=303). Delivered clear statistical interpretation for both technical and non-technical audiences.

RRStudioChi-SquareHypothesis TestingData Cleaning

Text Sentiment Classification

DIMENSIONALITY REDUCTION · DATA MINING

Preprocessed an 800-instance sentiment dataset using TF-IDF and Lovins stemming (1,199 features). Applied Random Projection to 900 features and evaluated the accuracy/efficiency trade-off across J48 and LibSVM classifiers — showing dimensionality reduction impact is algorithm-dependent.

WEKATF-IDFJ48LibSVMRandom ProjectionData Mining
Skills

Technical Stack

Languages & Libraries
  • Python★★★★★
  • SQL (MySQL)★★★★★
  • R / RStudio★★★★
  • Pandas / NumPy★★★★★
  • Matplotlib / Seaborn★★★★
  • Java★★★★★
ML & Statistics
  • Scikit-learn★★★★★
  • XGBoost / LightGBM★★★★★
  • MLflow★★★★
  • Feature Engineering / ETL★★★★
  • TensorFlow / PyTorch★★★
Cloud, BI & Tools
  • Power BI & DAX★★★★★
  • Excel (PivotTables · Power Pivot)★★★★
  • Azure App Service (Deployment)★★★★
  • Flask API / Gunicorn★★★★
  • Git / GitHub★★★★
  • WEKA · Tableau (familiar)★★★
About

Who I Am

I'm a data professional based in Hatfield, UK, holding an MSc in Data Science & Analytics earned with Distinction from the University of Hertfordshire. My BTech background in mechanical engineering gives me a structured, mathematical mindset I now apply to building ML systems that solve real problems.

I work across the full data lifecycle — from data cleaning and statistical analysis through to model deployment on Azure. I'm equally comfortable building predictive models in Python and delivering Power BI dashboards that non-technical stakeholders can act on immediately.

Actively seeking roles as a Data Scientist, Data Analyst, or BI Analyst. Right to work in the UK. Open to hybrid and on-site roles.

ENGINEERING BACKGROUND
BTech Mechanical Engineering — systems thinking and mathematical modelling applied directly to data science.
END-TO-END MLOPS
From MLflow experiment tracking to live Flask API on Azure App Service — full production awareness, not just notebooks.
STAKEHOLDER COMMUNICATION
Experienced at translating statistical outputs into clear, actionable recommendations for mixed audiences.
EDUCATION & CERTS
MSc Data Science & Analytics
University of Hertfordshire · 2023–2025
Machine Learning, Data Mining, Applied Data Science, Statistical Methods, Team Research & Development.
DISTINCTION
BTech Mechanical Engineering
Dhanekula Institute of Engg & Tech · 2018–2022
Analytical and mathematical foundation directly transferable to data science.
Software Testing
QSpiders, Bangalore
Manual Testing · Selenium · Java · Python · MySQL
COMPLETED
Contact

Let's Connect

I'm actively looking for Data Scientist, Data Analyst, and BI Analyst roles across the UK. Right to work — no sponsorship required. Reach out any time.