Call center services in india

Anybody tried Laguna S 2.1 (by Poolside)? 4 by spottedmarley | 0 comments on Hacker News. I just found out about it but I hadn't ever seen it mentioned and it sounds really interesting. Im pulling laguna-s-2.1:q4_K_M right now

Why Fireworks doesn't support Voice AI 3 by kushalpatil07 | 0 comments on Hacker News. I started thinking over why doesn't fireworks support voice models. There are really good opensource models available now, like parakeet, kokoro, Qwen ASR etc but no way to use it without managing a bunch of GPUs yourself. Even LLMs like Gemma 4 used by voice agents are not supported. Then I figured that the inference platform needs to be optimized differently for the kind of usecase you are using. Lets take an example for LLMs, not even STT and TTS. - Coding agents -> lot of cached input, needs to optimize for KV cache - Creation slides/blogs -> lots of output, needs to optimize for speculative decoding - Voice LLMs -> Cached input small output, not yet figured out on how to optimize this. So TTS and STT is a completely different ballgame. What I don't know is the timing, do people want to use open source models like kokoro, parakeet, Qwen etc RIGHT NOW?

Ask HN: Who Has This Pain? 3 by swk-phil | 2 comments on Hacker News. has to open media files from untrusted sources in sensitive environments as part of their daily workflow + fear that a zero-day exploit could be in one of those media files + cyber attack would have huge impact on the business

Ask HN: What was your big failure? How did you get around it? 2 by jspann | 0 comments on Hacker News. Sincerely a guy who took a gamble on the last few years and is about to lose

The LLMs Problems 2 by noreplydev | 1 comments on Hacker News. What problems are you all facing while building with LLMs?

Ask HN: Does AI research need "world models" more than bigger LLMs? 2 by unjuno | 1 comments on Hacker News. It seems that generating hypotheses is no longer the main problem. The harder problem may be continuously collecting real-world data, building accurate world models, and using them to guide experiments. Is this where automated research is heading?