
Prashanth Rao
@tech_optimist
Followers
1K
Following
13K
Media
113
Statuses
3K
AI engineer working with graphs & LLMs @kuzudb. Blogging @ https://t.co/gLektr01zQ
Toronto 🇨🇦
Joined October 2013
There are too many "hello world" examples of DSPy out there, i.e. folks just using off-the-shelf modules like Predict and CoT. YouTube is littered with them (not very helpful to build real projects). I'd like to do more to share better examples, but hope to see others to do too!.
@tech_optimist @MaximeRivest Yes definitely you should be building custom Modules. All DSPy modules are implemented in DSPy. They’re not special at all.
0
0
2
Seems like it 😭.
@DSPyOSS Every time I think I know what DSPy is I see a new DSPy tweet and realize I have no idea what DSPy is 😭.
0
0
0
Great summary of why SWE-bench is flawed, and more generally, why all benchmarks are flawed. Always build your own eval suite!.
SWE-bench Verified is one of the main benchmarks to assess AI coding skills. But what does it actually measure?. We found that it's one of the best tests of AI coding, but limited by its focus on simple bug fixes in familiar repositories. Here’s a summary of our article 🧵
0
1
3
Can't wait to get *deep* into the weeds with this thing! The initial code is already looking good! Looking forward to making more content on this in the future for those who cannot attend. But if you have the budget to do so, buy your ticket now!.
Join our AI engineer @tech_optimist and many other builders at the @_odsc Agentic AI summit (online) in July! We'll be covering how to build intelligent agent workflows powered by Graph RAG and production-ready knowledge graphs. Registration link here:.
0
0
3
RT @0xblacklight: I built an open-source deep research agent with Gemini, @ExaAILabs and @boundaryML's BAML + a terminal UI. (also has a s….
0
7
0
As they say, those first thousand followers take the longest to arrive! Looking forward to seeing where we can go from here 😁🚀.
Thank you for following our journey and helping us reach 1k followers here on X! The next several months are incredibly exciting for @kuzudb, and there will be loads more updates and great new features. Please spread the word, and we appreciate your continued usage of Kuzu!.
0
0
3
Such amazing news!!! Slowly but surely, @lancedb is building one of the most cracked Rust 🦀 dev teams on the planet. All while giving back to the open source ecosystem. Check out @OnlyXuanwo's work and use Lance! 🫶🏽.
Grateful waves goodbye, curious waves hello. See you at the next PR review — perhaps in @lancedb , perhaps back in @DatabendLabs, always in open source.
2
0
5
Thank you for making this @bllchmbrs! This is exactly the sort of educational content that makes open source so amazing and why AI is just exploding in all domains. Let's continue giving back in whatever way we can. Will be deep-diving into this in due course!.
so it took a bit more than 20 days but for those just getting started with @DSPyOSS . I go through all of the fundamentals of DSPy for 0 to hero. All code is visible so even if you don't have time to go through a full tutorial you can get a sense for the.
0
0
3
This is amazing! I was trying to squeeze in the time to work on something similar, but @arundsharma beat me to it. Great work!!! Looking forward to dive into the code, and @DSPyOSS in general 🤝🏽.
0
2
9
More exciting new algorithms on the way in @kuzudb in upcoming releases! This is thanks to some great work by @sdht0 and the Kuzu team. In the meantime, check out the docs here:.
Native, on-disk, scalable, efficient graph algorithms in an open source graph database? Sign me up! Can't wait to see people run these algorithms on *really* large graphs with @kuzudb!.
0
1
5
RT @doesdatmaksense: DSPy this, DSPy that.it has taken all over my timeline. wait brother, give me a weekend with some free time, i will tr….
0
2
0
Happy to have published this first post (of hopefully many more) on building towards agentic flows in Graph RAG. Rather than using an agent framework right off the bat, each of which have their own abstractions, I worked with pure BAML prompts and @kuzudb as the graph database.👇🏽.
We just published a new blog post on Graph RAG with router agents in @kuzudb that can call vector search tools via @boundaryML #BAML prompts. 👇🏽. Read the thread below to see the summary! 🧵. 1/9
1
2
8
For context:.
Today we’re announcing our $30 million Series A. This round is led by @Theoryvc with support from @CRV , @ycombinator, @databricks, @runwayml , @ZeroPrimeVC , @swift_vc, and more. Your belief in a future powered by multimodal data brings us one step closer to that reality.
0
0
0
TIL that you can have brilliant, crazy ideas that build successful companies, over capsaicin-induced meals. So happy for @changhiskhan and the entire @lancedb team, it's been quite a ride for me cheering these guys on too!
1
0
4