Nick Ryan
Data Scientist @ Microsoft for Startups | Enabling the Next Generation of Founders | Building Generative AI Solutions at Scale | TCS
- Role
- Data Scientist at Microsoft
- Location
- Redmond, WA, US
- LinkedIn followers
- 500 followers
About Nick Ryan
4 years, 50+ bespoke AI solutions, and 150k startups analyzed… or so goes my Microsoft…
Experience
Data Scientist
Jan 2022 — Present · US
Microsoft-for-Startups Projects- An LTV model to predict Azure revenue for each startup from early indicators (R^2 = 0.8)- A COGS model to anticipate per-startup credit usage based on early indicators (R^2 = 0.8)- A recruitment model to identify high-potential startups from the broader startup ecosystem with high potential for Azure spend (R^2 = 0.95)- A churn model to determine which startups would leave the MfS program and convert-to-paying Azure customers (R^2 = 0.92) versus those who churned from Azure completely (R^2 = 0.87)- A Natural Language to SQL + Matplotlib Agent for internal use to enable 100+ non-data Microsoft-for-Startups team members to get insights from our database on-the-spot - An agentic web scraper tool to sift through news articles and blog posts for mentions of recent investments in startupsAI Co-Innovation Labs Projects- A rapid-responding 911 dispatch Co-Pilot supplying cue cards for on-call 911 dispatchers with information from graphRAG knowledge bases of car owners manuals, standard operating procedures, cctv footage, etc… - A Kawasaki AS Programming agent taking environment cues from vision equipment to write Assembly Script executable in environments the robot had never seen before - A marketplace management tool to consolidate multiple sellers of a single product under one listing in the marketplace database, monitoring for information accuracy and duplicate prevention - A conversational interface for a grocery store’s inventory enabling shoppers to receive product recommendations and build recipes with in-stock items, and floor workers to receive insights about front-of-house and back-of-house stock-outs - A natural-language-to-WQL (Workday Query Language) agent and a natural-language-to-SoQL agent, back when neither query language was captured in GPT3’s training data - A military simulation agent to help add/remove/reposition entities across maps to aide in military training exercises - and many, many more…
Education
UC Santa Barbara
Bachelor’s Degree, Statistics and Data Science
2017
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.