- 👋 Hi, I’m @unmeshmali25
- 👀 ML Engineer at CVS Health
- Working at inference scale of 80 million customers per week.
- Working on
- @CVS : Recomendation Engine, Offer engine, MLOps infrastructure overhaul.
- A full-stack retail app with AR-based product identification and coupon matching (React, Supabase, Python/Railway)
- An AI agent orchestration framework — 372 LLM-powered personas simulating real shopper behavior over 3+ synthetic years (LangGraph, LangSmith, asyncio)
- A complete MLOps platform for recommendations, offer optimization, segmentation, and CLV modeling (Terraform, GCP, DVC, MLflow, GitHub Actions)
- 📫 You can reach me at unmeshmali@gmail.com or send me a connection request at https://www.linkedin.com/in/unmesh-mali/
- San Francisco
- https://www.linkedin.com/in/unmesh-mali-92596799/
Pinned Loading
-
tfranza/ibotta_fraud
tfranza/ibotta_fraud PublicAn ML product that make use of an unsupervised anomaly detection model for Fraud Detection using the Berka 'financial' dataset.
Jupyter Notebook 3
-
Automating-Workloads-Using-ARM-Template
Automating-Workloads-Using-ARM-Template PublicAutomating Workloads Using Azure ARM Template for a fictional Rand Enterprise Corporation
-
aws-devops-cicd-yoda-insurance-microservice
aws-devops-cicd-yoda-insurance-microservice PublicPython
-
Azure-CLI-Distribute-Incoming_traffic
Azure-CLI-Distribute-Incoming_traffic PublicCreate Highly Available Architecture by Distributing Incoming Traffic among Healthy Service Instances in Cloud Services or Virtual Machines in a Load Balanced Set with the Help of Command-Line Inte…
-
Building_and_Evaluating_Advanced_RAG
Building_and_Evaluating_Advanced_RAG PublicThe repository uses llamaindex to build 3 levels of RAG applications on my personal blog. It also uses TrueLens to evaluate each of the RAG methodologies based on context relevance, answer relevanc…
Jupyter Notebook 1
-
serverlessLLM-AWS-bedrock
serverlessLLM-AWS-bedrock PublicBuilding an audio summarization app using LLM from Bedrock and using AWS Lambda to make it serverless
Jupyter Notebook
If the problem persists, check the GitHub status page or contact support.

