Mehar Singh Bawa

ML / AI Engineer  ·  Data Science

Mehar Singh Bawa

I design and deploy agentic AI systems, from multi-agent orchestration pipelines & production-ready ML models to cloud-native applications, drawing on 6 years across cloud data engineering and applied machine learning.

Projects

Systems that turn data and AI into decisions.

Agentic AI Resume Screening System dashboard
Agentic AI · Multi-Agent Orchestration

Agentic AI Resume Screening System

A four-agent pipeline that screens resumes against job descriptions, with a dedicated Critic Agent running a bounded reflection loop to catch and correct weak evaluations before final output.

  • LangGraph
  • LangChain
  • LangSmith
  • OpenAI API
  • Pydantic
  • Reflection Pattern
  • Streamlit UI
  • Python
Financial Sentiment Analysis using FinBERT
NLP · Fine-Tuning

Financial Sentiment Analysis using FinBERT

Fine-tuned BERT for financial sentiment classification on Financial PhraseBank, trained end-to-end on a single T4 GPU.

  • PyTorch
  • HuggingFace
  • Text Classification
  • Transformer Fine-Tuning
  • NLP
  • 96.5% F1
Retail Customer Analytics and Sales Forecasting dashboard
Analytics · Forecasting

Retail Customer Analytics and Sales Forecasting

Customer segmentation and demand forecasting built on RFM analysis, Auto-ARIMA time-series models, and Apriori-based market basket analysis.

  • RFM Segmentation
  • Auto-ARIMA
  • Apriori
  • Association Rule Mining
  • FP-Growth
  • Recommendation System
  • Python
Diabetes Prediction Project
ML Systems · MLOps

Diabetes Prediction Project

A production-ready prediction service with a served model, experiment tracking, and a relational store for inference logging.

  • FastAPI
  • MLflow
  • PostgreSQL
  • Boosting Algorithms
  • Drift Detection
  • MLOps
  • Python
Airline Passenger Satisfaction Prediction
Classification · Predictive Modeling

Airline Passenger Satisfaction Prediction

A classification model predicting passenger satisfaction from flight and service-quality features, with model comparison and evaluation.

  • Python
  • pandas
  • scikit-learn
  • Classification
Fraud Detection Analytics using Snowflake
Analytics Engineering · SQL

Fraud Detection Analytics using Snowflake

A CTE-based composite risk scoring model built in Snowflake, framed in feature-engineering terms for fraud pattern detection over transactional data.

  • Snowflake
  • SQL
  • Risk Scoring
Electric Vehicle Data Analysis using Power BI
BI · Data Visualization

Electric Vehicle Data Analysis using Power BI

Data analysis on EV adoption and sales trends, charging efficiency, and environmental impact, delivered through interactive Power BI dashboards for data-driven decision-making.

  • Power BI
  • DAX
  • Data Visualization
  • Sales Performance
  • EDA
Telco Customer Churn Model Comparison
Classification · Model Comparison

Telco Customer Churn Model Comparison

A benchmark of multiple classification algorithms for predicting customer churn, compared across accuracy, precision, and recall.

  • Python
  • scikit-learn
  • EDA
  • Model Benchmarking
SAP ERP Simulation Game
ERP · Business Simulation

SAP ERP Simulation Game

A hands-on SAP ERP simulation covering procure-to-pay and order-to-cash business processes in a live enterprise system.

  • SAP ERP
  • Power BI
  • Business Process
Linear Regression Analysis in Python
Statistics · Regression

Linear Regression Analysis in Python

A from-the-ground-up walkthrough of linear regression, covering assumptions, diagnostics, and model evaluation.

  • Python
  • statsmodels
  • Regression

Skills

Agentic AI & LLM

  • LangGraph / Multi-Agent Systems
  • RAG Pipelines
  • Prompt Engineering
  • OpenAI / Groq APIs

ML & Data Science

  • PyTorch / HuggingFace
  • Time-Series Forecasting
  • Ensemble Modeling
  • Feature Engineering

Cloud & Data Infrastructure

  • AWS (Certified Developer Associate)
  • Terraform / CI-CD / ECS
  • Snowflake / PostgreSQL
  • Vector Databases

Languages & Tools

  • Python / SQL
  • Java / Spring Boot
  • FastAPI / MLflow
  • Git / Docker

Experience

Graduate Research Assistant Colorado State University

Building multi-agent pipelines, RAG systems, and LLM-based classification tools under faculty research supervision, alongside an M.S. in Computer Information Systems.

Data & Cloud Consultant Capgemini · 5 years

Cloud infrastructure, data engineering, and client engagement, partnering directly with stakeholders on key technical decisions. Delivered CI/CD automation, API development, application migrations, and large-scale financial data pipeline improvements.

Bachelor's in Computer Science & Engineering SRM University · 2019

Foundational coursework in data structures, algorithms, databases, and software engineering.

Let's build something together?

Open to AI/ML Engineer and Data Scientist roles.