Skip to content
View Bhagyesh1207's full-sized avatar

Highlights

  • Pro

Block or report Bhagyesh1207

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Bhagyesh1207/README.md

Hi, I'm Bhagyesh Patel πŸ‘‹

Data Analyst | Data Quality & AI

Portfolio LinkedIn Email

πŸ“ Toronto, ON Β Β·Β  πŸŽ“ Sheridan College, Computer Science (Data Analytics) Β Β·Β  πŸ“„ IEEE published

I build data pipelines people can trust: validation rules that catch bad data early, governance that keeps models accountable, and LLM workflows that are tested like any other system.

πŸ’Ό Now

Data Analyst, Red Pico (Mar 2026 to present)

  • βœ… Built automated validation rules in Informatica across 90% of critical pipelines in one quarter
  • 🧭 Governance workflows in Dataiku for model lineage, bias metrics and compliance
  • ⚑ SQL tuning and transformations on 100k+ row tables for internal dashboards

πŸ•˜ Before

  • πŸ€– Generative AI Associate, Innodata Canada (Aug 2025 to Mar 2026): training data and human-in-the-loop evaluation that improved LLM accuracy, fluency and coherence by 25%+
  • 🚚 AI/ML Developer, Sheridan Centre for Applied AI (Sep 2024 to Aug 2025): fleet KPI extraction with a locally deployed Llama 3, 60% better extraction accuracy through prompt engineering, BERTopic taxonomy work
  • 🩺 Machine Learning Data Scientist, Sheridan Centre for Applied AI (May 2024 to Aug 2024): 100,000+ patient records, 40% accuracy gain with a voting ensemble

πŸš€ Featured work

project what it is
πŸ›‘οΈ DQ Rules Agent Profiles a dataset and generates explainable data quality rules from its own statistics, no LLM needed. Exports to Great Expectations, pandera and SQL. ▢️ Live demo
πŸ† FairLens AI hiring platform with PII anonymization and a diversity dashboard. 2nd best capstone project award
πŸ“„ LLM-Summarize-BERTopic Research published at ACDSA 2025 (IEEE): LLM summaries improve BERTopic topic modeling. Read the paper
πŸ§ͺ Automated KPI Validation Tests LLM extracted KPIs for faithfulness to the source document with DeepEval
πŸ’¬ AWS RAG Chatbot Document-aware chatbot on AWS Bedrock, Titan embeddings and OpenSearch

πŸ› οΈ Tools

Tech stack

πŸ“Š Data: SQL, Python (pandas, NumPy), Informatica, Alation, Dataiku, Power BI, Spark, Hive, MongoDB

πŸ›‘οΈ Data Quality & Data Governance: validation rules, data profiling, lineage, bias metrics and compliance workflows (Dataiku, Informatica, Alation, Great Expectations, pandera)

🧠 ML & AI: scikit-learn, PyTorch, TensorFlow, LLMs, RAG, BERTopic, LangGraph, DeepEval

☁️ Cloud & apps: AWS (S3, Bedrock, OpenSearch, Amplify, CloudWatch), Google Cloud, Flask, FastAPI, Dash

πŸŽ“ Education

Honours Bachelor of Computer Science (Data Analytics), Sheridan College, 2025 Β· GPA 3.55/4

Pinned Loading

  1. Fairlens Fairlens Public

    FairLens is an AI-powered hiring platform that reduces bias in recruitment by combining machine learning, large language models (LLMs), and real-time diversity analytics. It supports resume parsing…

    Python

  2. Automated-KPI-Validation-with-Deep-Eval Automated-KPI-Validation-with-Deep-Eval Public

    Tests that KPIs an LLM extracts from a document are faithful to the source, using DeepEval (pytest).

    Python

  3. AWS-RAG-Based-Chatbot AWS-RAG-Based-Chatbot Public

    Document-aware RAG chatbot on AWS Bedrock (Claude v2), Titan embeddings and OpenSearch.

  4. dq-rules-agent dq-rules-agent Public

    Profile a dataset, get explainable data quality rules from its own statistics (no LLM), export to Great Expectations, pandera or SQL.

    Python

  5. Fleet-KPI-Weight-Agent Fleet-KPI-Weight-Agent Public

    Jupyter Notebook

  6. LLM-Summarize-BERTopic LLM-Summarize-BERTopic Public

    Enhancing Topic Modeling via LLM-Driven Summarization