Open to opportunities

Valarmathi Sri

@valarmathisri

Message

Production-ready Data Scientist building ML and GenAI pipelines for measurable business impact.

United Kingdom

Message

What I'm looking for

I’m looking to build production-ready ML/GenAI solutions—end-to-end pipelines, forecasting and NLP, and stakeholder-ready dashboards—while working with teams that value practical impact, fast ownership, and clear data storytelling.

I’m a production-ready Data Scientist with 3+ years of hands-on experience delivering 15+ real-world projects across telecom, healthcare, finance, e-commerce, and digital media. I build end-to-end ML pipelines, integrate outputs into Power BI dashboards, and turn insights into strategic, stakeholder-ready decisions—grounded in supervised learning, unsupervised learning, time series forecasting, and deep learning.

Recently, I built VM JD Matcher, a GenAI-powered FastAPI + Streamlit application that uses Hugging Face transformer embeddings (all-MiniLM-L6-v2), regex skill extraction, and semantic similarity scoring to generate resume–job description match percentages, overlapping/missing skills analysis, and tailored guidance for both jobseekers and recruiters.

I also develop practical AI systems like RAGFlow (a PyPI-published Python library for Retrieval-Augmented Generation pipelines with FAISS-based vector search) and RAGFlow/RAG applications in Streamlit. In applied work, I’ve delivered forecasting & inventory intelligence, LSTM-based financial time-series prototypes, and robust model-selection workflows using temporal validation, aiming for fast execution, ownership, and dependable results.

Experience

Work history, roles, and key accomplishments

Maths TA / Cover Supervisor

Multiple Schools Across Hertfordshire

Jan 2017 - Jan 2025 (8 years)

Delivered mathematics lessons across KS3–KS5, planning lessons and assessments, supporting GCSE preparation, and coordinating with teachers and parents to address student learning needs.

Lesson Planning Assessment Marking Student Mentoring Classroom Management

Junior Software Developer – UKPC

UKPC

Oct 2022 - Dec 2022 (2 months)

Built Python automation scripts for converting and extracting content, configured annotation tooling, and supported SharePoint Online updates with branding and PowerShell automation, including Power BI reporting.

Python Data Cleaning Whitespace Prodigy Annotation Tool HTML CSS PowerShell SharePoint Online Power BI

Junior Software Developer – Base2

Base2

Apr 2022 - Jun 2022 (2 months)

Developed applications using ASP.NET Core/MVC and JavaScript, supported SharePoint administration with PowerShell, improved search performance in SharePoint lists, and delivered reporting using Power BI.

ASP.NET Core MVC C#JavaScript HTML5 CSS SQL Server PowerShell SharePoint Administration SPServices jQuery Excel Power BI Data Analysis data preprocessing Feature Extraction Performance Optimization

Data Scientist – Techjays

Techjays

Jan 2026 - Present (6 months)

Built a forecasting and inventory intelligence system for 80 SKUs, delivering demand predictions, safety stock, reorder points, and SKU-wise model selection using time-series validation.

Time Series Inventory Optimization Feature Engineering ARIMA Holt Winters XGBoost Automated Model Selection Reorder Points Demand Forecasting

Data Scientist – Intelligence Tech

Intelligence Tech Solution

Nov 2025 - Present (8 months)

Designed a trading bot prototype using LSTM-based financial time-series forecasting with walk-forward validation and reproducible pipeline outputs for backtesting and trading simulations.

Financial Forecasting Deep Learning Dropout Regularization Gradient Clipping Walk Forward Validation Paper Trading Live Trading

Data Scientist – VM JD Matcher

VM JD Matcher

Jan 2025 - Present (1 year 6 months)

Designed and implemented a GenAI-powered application to score resume–job description alignment using transformer embeddings and regex-based skill extraction, with an interactive Streamlit UI and FastAPI endpoints.

Embeddings Hugging Face (All MiniLM L6 V2)Regex Skill Extraction Resume JD Matching Streamlit fastAPI Deployment

Data Scientist – Unified Mentor

Unified Mentor

Worked on attrition intelligence using classification models and delivered Power BI dashboards and app analytics.

Classification Logistic Regression Random Forest SVM Power BI App Analytics

Data Scientist – Envision Virtue

Envision Virtue

Built salary prediction regression pipelines and developed climate sentiment modeling using NLP and unsupervised techniques.

regression Ridge Lasso NLP TF IDF Topic modeling Sentiment Analysis

RAG Python Library

Ragflowvm

Designed and developed a lightweight, modular Python library for building Retrieval-Augmented Generation (RAG) pipelines, including document ingestion, chunking, embedding generation, and FAISS-based vector search.

RAG Python Faiss Vector Search PyPI

Document-Grounded Q&A System

RAGFlow

Built an end-to-end document-grounded question-answering RAG application with a complete retrieval and generation pipeline and a Streamlit interface.

RAG Streamlit Large Language Models Evidence Based Question Answering

Data Analyst – Uptrail

Uptrail

Delivered multiple analytics projects, building classification and regression models and creating Power BI dashboards with advanced DAX and automated ETL workflows.