MS CS grad trying to pivot into AI inference and ML infrastructure, and honestly I'm a bit lost.
I'm currently learning transformer internals and vLLM serving, and I'm planning to study computer architecture next. There still feels like there's so much to learn, and I'm not sure what to prioritize.
If you work in ML infra, what would you focus on if you were starting today? I'd really appreciate any advice.
#MLInfra #Inference #LLM #AIInfrastructure #MLOps #DistributedSystems #PyTorch #vLLM #AIEngineering #NewGrad
@AravSrinivas I Noticed Perplexity returning raw XML traces like
<xai:function_call name="search_web">...</xai:function_call>
instead of a formatted response today. It's giving out function call + query params instead of intended output.
The Comet waitlist is expanding too fast despite all the daily invites. So, weโre going to increase our daily Comet invites going forward. Starting tomorrow.
Huge thanks to all the open source projects that've made a lot of the tech we rely on in the world possible:
Linux
Git
FFmpeg
PyTorch & TensorFlow
Apache & Nginx
MySQL, PostgreSQL, SQLite
Chromium & Firefox
GCC & LLVM
Docker & Kubernetes
Also, all the open-weight LLMs... and all the programming language interpreters, compilers, and frameworks.
I can list 100+ more "niche" open source projects I've used & loved (e.g., OpenCV โค๏ธ). But I just wanted to give a continued big thank you to the open source community for everything you do for humanity ๐
Working on an open-source schema crawler + query engine.
It connects to your database, learns the schema, and lets you retrieve data in plain English.
Over time, it adapts:
caches common logic
improves queries via feedback
no raw data needed, just schema & metadata
Useful for enterprises especially in supply chain (I've felt the need๐)
Looking for collaborators. DM me
Working on an open-source schema crawler + query engine.
It connects to your database, learns the schema, and lets you retrieve data in plain English.
Over time, it adapts:
caches common logic
improves queries via feedback
no raw data needed, just schema & metadata
Useful for enterprises especially in supply chain (I've felt the need๐)
Looking for collaborators. DM me
Working on an open-source schema crawler + query engine.
It connects to your database, learns the schema, and lets you retrieve data in plain English.
Over time, it adapts:
caches common logic
improves queries via feedback
no raw data needed, just schema & metadata
Useful for enterprises especially in supply chain (I've felt the need๐)
Looking for collaborators. DM me
@AravSrinivas
Perplexity is great; being able to use any LLM I want in one place is cool.
Just wish it followed the conversation thread better. I asked a follow-up on the Windsurf saga from the discover tab, and it started explaining windsurfing ๐ Even in web searches, it often loses track when I ask follow-ups.
@perplexity_ai ..