Cerebrium

We're excited to announce that we're part of the Y Combinator W22 batch! Sign up for early access here: https://www.cerebrium.ai/ Y Combinator is the most renowned startup accelerator in the world and has backed companies like Airbnb, Stripe, Coinbase and Reddit, Inc. and we’re honored to be joining this community of founders. We would like to thank our group partners, Tim Brady, Dalton Caldwell, Harj Taggar, and Tom Blomfield for helping us on this journey as well as Charl Jacobs and our customers that continue to believe in us and have shown us continued support. Please follow us as we work to enable businesses to be data and AI-driven without a single line of code or an entire data team. Additionally, if you want to be a part of something special, check out our careers page below (link in comments)

Gary Vaynerchuk mentioned that in the future users would be able to shop the items in movies and videos in real-time - well that day is today! We built a showcase demoing how we can detect items in a video, search an entire product catalog (or the internet) and give you a link to go buy that very item, all in <500ms. With a lot of content moving towards video, this is just an example of how companies can extract data and provide value to the end user in real-time! If you want to give the basic live demo a try: https://lnkd.in/ey2RRkwJ You can read more about how we achieved this here: https://lnkd.in/ehb8dEmb #ecommerce #ai #streaming #startup

Gary Vaynerchuk mentioned that in the future users would be able to shop the items in movies and videos in real-time - well that day is today! We built a showcase demoing how we can detect items in a video, search an entire product catalog (or the internet) and give you a link to go buy that very item, all in <500ms. With a lot of content moving towards video, this is just an example of how companies can extract data and provide value to the end user in real-time! If you want to give the basic live demo a try: https://lnkd.in/ey2RRkwJ You can read more about how we achieved this here: https://lnkd.in/ehb8dEmb #ecommerce #ai #streaming #startup

We created an demo application where you can ask Andrej Karpathy, a well-known figure in the AI world, questions in real-time based on his videos explaining complex AI concepts. Personalised tutoring is here! Demo: https://lnkd.in/dp-5PyK3 Blog: https://lnkd.in/d5y9guac Code: https://lnkd.in/dPdGhnYs. Try it out and let us know what you think! If you would like to build something similar - please reach out! #generativeai #education #aiagent

🔔 Exciting Updates to the Cerebrium Platform! 🚀 We’re thrilled to share some exciting updates to the Cerebrium platform, released over the past few weeks. Based on your feedback, we’ve been focusing purely on performance improvements! Here’s what’s new: 1. Faster Cold Starts We’ve made a significant change to our runtime, enabling cold starts of approximately 3 seconds. This means improved response times and a better overall user experience for your customers. 2. Improved Build Times Builds are now 60% faster! First-time builds take around 2 minutes, and subsequent builds are down to roughly 10-20 seconds. This ensures you get valuable feedback sooner and can deploy your models to production much quicker. 3. Updated Website Our website has been revamped to reflect the changes we’ve made over the past year. We invite you to check it out and share your feedback with us! 4. Starter Templates To help new users get up and running with their use case faster, we’ve added new starter templates. If you have suggestions for other templates, please let us know! Website: https://www.cerebrium.ai #ai #serverless #infrastructure #gpu

Voice is going to continue to be a bigger and bigger part of how we all interact with generative AI. But one thing that's hard, today, is building voice bots that respond as quickly as humans do in conversation in a way that feels natural. Thats why today, in partnership with Daily and Deepgram we released the fastest AI voice agent with voice-to-voice response times of ~500ms. Companies are now able to launch a real time voice AI agent at a fraction of the cost and engineering effort, unlocking so many real tangible use cases for businesses and startups of all sizes. We kept engineers in mind allowing you to swap in and out various components of the pipeline. Code is available online! Let us know what you think! Demo: https://lnkd.in/ddUYFYtD Docs: https://lnkd.in/d6XMbc2f #generativeai #aiagents #startups #voiceai hashtag #generativeai hashtag #aiagent hashtag #ai hashtag #startup hashtag #engineering

We are proud to announce our new partnership with Decart AI, allowing our customers to process 1 million tokens of Llama 2 70B for just $0.50! Decart built a proprietary LLM inference engine from scratch in C++ and CUDA to outperform other existing engines (vLLM, TGI, etc.). The engine is hosted on the Cerebrium platform, making it accessible to developers with <5s cold start times. We are hoping this partnership will create a better user experience with lower latencies while at the same time making it more affordable for companies to run these models at scale. Try out the API here: https://lnkd.in/en9yiyqZ Read the official Blog post here: https://lnkd.in/eXzC4htK #generativeai #openai #llama2 #nvidia Currently supported models: Llama 2 70B, Mistral 7B and Zephyr 7B.

We have partnered with Daily to support engineers building real-time apps that combine voice and video with ML/AI without worrying about any of the infrastructure.

Great article by Lightspeed on how companies will evolve in their MLOps Architecture as their solutions get more mature. The models offered by OpenAI and Anthropic while very generalist, are too expensive for companies to offer in production. Besides, the latency is another issue especially with multi-modal implementations or companies using agents. However, the infrastructure to support alternative implementations hasn't really existed without significant effort and experimentation. If you are interested to see how we have solved this issue and helped companies reduce their OpenAI bill by 50%+ at Cerebrium get in touch. https://lnkd.in/e_EddXRq #startups #machinelearning #openai #mlops