#pydatabos last but definitely not least Mike Woodward taking us through his trials and tribulations with using cursor for code. Seems it's not quite ready for DS work!

#pydatabos @shirandudy.bsky.social on bias in LLM recommendations docs.google.com/presentation...

#pydatabos @allendowney.bsky.social's super interesting framing - teaching linear algebra with math + code rather than math THEN code docs.google.com/presentation...

#pydatabos a peek at Center for Effective Philanthropy's stack for processing survey data @phenomanaal.bsky.social

#pydatabos our organizer Bryan Romas kicks us off with a "build your own git" coding exercise. Check it out github.com/Bcromas/buil...

ODSC ‘25 Recap Conference recap time! I always love putting these together, I feel like it helps me solidify my understanding of the material and gives me something to refer back to when I think of one of the talks ...

Put together a wrap up for #odsc! Materials and some notes from a bunch of great talks! bpben.github.io/odsc_25_recap/

#aitinkerers SPARC orchestrates different agents with different roles to automate code writing

#aitinkerers vibe coding done right take a nap and let it run

#aitinkerers Coach Sensai project - an agile and delivery coach. Integrates nicely with slack and helps manage tickets and stand ups

#aitinkerers docrouter.ai example with generated rubric and scoring by llm

#aitinkerers using AI for partnerships with influencers: apify to get structured data and browser use agent for getting info like reel views

AI Tinkerers - Boston AI Tinkerers - Boston - Get updates and be the first to know when we launch events in Boston - 30 posts

#aitinkerers At AI Tinkerers Boston boston.aitinkerers.org listening to some interesting talks on AI projects

#odsceast how to implement QA for testing agents

#odsceast the great Serg Masis on the importance of QA processes

#odsceast some metrics for reliability for agents that are not part of current benchmarks @randomwalker.bsky.social

#odsceast @randomwalker.bsky.social discussing the limitations of agent benchmarks arxiv.org/abs/2407.01502

#odsceast a realistic take on AI agents with @randomwalker.bsky.social

#odsceast Boltz-1 open source reimplementation of alphafold3

#odsceast keynote speaker Regina Barzilay talking about the importance and limitations of alphafold3

#odsceast algorithms for fine tuning LLMs GitHub.com/SethhWeidman/odsc_east_2025_grpo_llama

#odsceast big punchline - LLM recommendations are biased towards wealthy and highly educated regions, disadvantaging other regions @shirandudy.bsky.social

#odsceast last but definitely not least @shirandudy.bsky.social on geographical bias in LLMs arxiv.org/abs/2504.05325

#odsceast next up data literacy and responsible AI

#odsceast quick hit sessions about AI system adoption. First up sustainable data science adoption

MetaSnake Products

#odsceast CatBoost with Matt Harrison store.metasnake.com love to see more traditional ML on this track store.metasnake.com

#odsceast last talk of the day on using LLMs for text classification and hybrid labeling

#odsceast Brian Lucena on uncertainty quantification. Excellent speaker always puts together very detailed info

#odsceast Pydantic.ai library for building agents. Love this setup - enforcing clear structure to the agent workflow

#odsceast learning about adaptive architecture in data engineering. Case study - S3! Works most of the time but is always updating and evolving.

Google Colab

Love this example - trusting by party affiliation colab.research.google.com/github/Allen...

#odsceast @allendowney.bsky.social