7/
Engineers became the bottleneck when analyzing sensor data. @movedot_ built AI agents that fix that, working hand in hand to analyze telemetry, video & documentation 100x faster. Starting with race cars
Congrats @FincoBruno & @g_movedot_ai
https://t.co/MtF2LemjDP
@conor_ai @iamfaranna@hyperspell Looking forward to the post when you will also take your first flight to implement it at the customer's site! First of many
NEWS | 🇺🇸 HMD Motorsports has teamed up with MOVEdot, an AI startup that will provide the @INDYNXT team with AI-based tools.
MOVEdot will support HMD engineers with driver performance analysis, driving reports with embedded video, strategy analysis, setup optimization and more.
@tooluseai Can confirm it was a great episode. I'd definitely use/pay for a tool that could do the YT summaries including daily emailing. And the winner was not Mike... or Ty... it was the Gemini's roast for sure
@tooluseai I will! At our company we are plugging engineering tools into LLMs, instead of traditional/general purpose tools. I will share ideas when I feel they match the channel!
@afshawnl As a user of CadQuery, I'm always interested in hearing more about your perspective on open source CAD. It has indeed made my life a lot easier and allowed for a lot more automation compared to closed source CAD. Designing engineering systems through code is just amazing.
@zinyando I'll be following those, all are interesting!
Larger multi-agent architectures that work well, can assign the responsibilities between them properly and don't get stuck in loops would be great too ;) or simply an architecture using a manager in crewai that's robust.
@zinyando@pyautogen@mem0ai These are the topics I've been researching lately. Not sure if it makes sense for you to cover them. But I can certainly keep you posted on topics I think match your content. Thanks!
@zinyando@pyautogen@mem0ai Or different ways to create training data for different applications
- Prompting/structured data (inquires and answers)
- Unstructured data: Raw text/text books (and how to make the LLM learn it)
- Tool use data sets
@zinyando@pyautogen@mem0ai One particular topic of interest is understanding when to RAG, when to fine-tune, at what point to make the transition, what are the costs that we would expect. Or to see a comparison between a model using RAG vs a model fine-tuned.