Praveen Sridhar

Machine Learning Engineer building in the healthcare and insurance space.

Co-founder of @tinkerhub, a non-profit educational initiative.

[Praveen Sridhar @psbots]

Yahoo Pipes remains a fond memory for almost everyone who has used it

Building an updated version of the same, a data mashup tool with a few basic "non-agentic" AI nodes that can filter/summarise/synthesise relevant info from news and other web feeds

[Praveen Sridhar @psbots]

Opus 4.6 in @AmpCode being direct - "what you've described is not buildable in one day"

and then proceeds to build it in 5 minutes!

will probably take a few more years for models to understand that their time estimates are very different from ours...

amuses me every time!

[Praveen Sridhar @psbots]

@Sauers_ I like the term "AI pair programming", which is how @paulgauthier describes it on the Aider website https://t.co/sa0u3nhdhH

[Praveen Sridhar @psbots]

@tambo_ai n64

[Praveen Sridhar @psbots]

@CoenHees @shopdev_ai shopdev

[Praveen Sridhar @psbots]

@al3rez @AnthropicAI @boltdotnew @lovable MVP

[Praveen Sridhar @psbots]

With new LLM launches on a daily basis, @vinayakh highlights the importance of having a sound and structured process for model selection - extremely important tool to have for any AI Engineer

- at the AI community day by Azure
@dearazure_net #dearazure

[Praveen Sridhar @psbots]

A few thoughts on LLM agents -
https://t.co/jbxmhTlOl8

[Praveen Sridhar @psbots]

The real Iron Man

@ylecun at the Meta Build with AI Summit in Bangalore

Thank you for all your contributions to keeping AI research accessible through libraries like Pytorch and models like Llama and Segment Anything

#GenAI #MachineLearning #DeepLearning

[Praveen Sridhar @psbots]

@theRichieMagnum surely, will do

meanwhile, to get started, here's a nice compilation on Machine Learning case studies by @EvidentlyAI
https://t.co/1a3O6WPimo

There are a lot more blogs than this list, but this in itself is large enough to last any curious learner for many many days!

[Praveen Sridhar @psbots]

There's a huge amount of valuable insights on real world Machine Learning and AI use cases across the developer blogs of various companies

Many of these remain hidden behind the daily flood of new model launches

Trying to collate and distill some of these

[Praveen Sridhar @psbots]

Do check out their blog post for more details - https://t.co/FgPLytJWUw

[Praveen Sridhar @psbots]

Goldman Sachs has an interesting post on how they used basic NLP and Machine Learning to improve their software development process. They wanted their engineers to succinctly describe change/merge requests in the codebase with both - “what” was being changed and “why”.

[Praveen Sridhar @psbots]

This use case also hints at the importance of the human element in adopting Machine Learning. Both the advocacy on the need for purposeful articulation of software changes and the tool that nudged teams to the desired outcome were essential for the success of the solution.

[Praveen Sridhar @psbots]

@1littlecoder Thanks!

[Praveen Sridhar @psbots]

Hacked together a browser based code interpreter for Claude 3.5 Sonnet as part of the Build with Claude contest

Its a browser extension that generates code for the given query using Claude API and uses PyScript, Pyodide and WASM to run full-fledged Python code within the browser

[Praveen Sridhar @psbots]

@subhashMkashyap