Data Science at Microsoft

Why is it so hard to ship a simple LLM feature?

Open AI Code Interpreter and GitHub Copilot X: Ideal partners?

Many facets of un-truth: LLM hallucination 101

With the increased adoption of LLMs across industries, hallucination has become…

Taking a hackathon project to production: Some lessons learned

How to leverage permutation tests and bootstrap tests for baselining your Machine Learning models

Fundamentals of building with LLMs: Question & answer on any document with ChatGPT in 30 lines of code!

The era of Co-Pilot product teams

Using multi-agent prompt engineering to fuel future product…

Azure AutoML: Quickly build high quality ML models

Evaluate combinations, select best performing ones…

Decoding the customer journey with graph node embeddings

How to use graph networks and algorithms to…

You ask the questions, GPT digs the insights

A Gen AI approach to business data analysis

From reviews to insights: Building analytic applications with Large Language Models in e-commerce

Categorically: Don’t explode — encode!

Read this article before you start preprocessing your data for tree-based models!

Data-driven injury prevention in baseball: Maximizing player performance and longevity

Creating a metadata graph structure for in-memory optimization

What they are, key components…

How to think about the future of AI, data science — and humans

Generative AI, OpenAI, and ChatGPT: What are they?

Live Site Incident escalation forecast

Machine Learning models for forecasting high impact incidents…

Modeling customer behavior with Markov chains

It’s natural for customers to differ in their consumption patterns and…

Using graphs to model and analyze the customer journey

A conceptual guide to getting started

Reasoning about change: Lessons learned from building a near real-time system for Azure pricing

Turning our API hackathon into an end-to-end data science portfolio project

Causal inference using synthetic controls

How to estimate causal effects using Machine Learning…