Pipeshift (YC S24)

And it’s a wrap for YC S24! Past 3 months have been nothing less than life-changing. Here are some highlights: - Our revenue has doubled (6-figures) - We went from <5 to 30+ customers - Scaled our waitlist to 200+ teams - Opened private access to our beta - Processed 4B+ tokens in last 30 days And, made friends (more like family) for life! Thank you Dalton Caldwell and Tyler Bosmeny for all the advice, feedback, and for being the best Group Partners anyone could’ve asked for. Even though the batch ended, YC doesn’t. It’s for life! Cause Pipeshift (YC S24) is just getting started. Enrique Pranav

Nothing is more wholesome than this! Public access of Beta coming by the end of 2024 :) Beta waitlist: https://tally.so/r/w4jxoY Check out pipeshift.com

Any company affected by the shut-down of OctoAI, don’t worry, we’ve got you covered! Pipeshift (YC S24) is offering FREE white-glove services to help you port without dismantling any of your LLM workloads. Book a call now: https://lnkd.in/gMra4-4X

?? Llama 3.2 just dropped. Open-source AI FTW! There are 4 variants: 1B, 3B, 11B, and 90B Llama 3.2 11B & 90B vision models deliver performance competitive with leading closed models — and can be used as drop-in replacements for Llama 3.1 8B and 70B. If you want to get Llama 3.2 up and running in production in your cloud at 20X better inference performance than Bedrock and Vertex AI (yes, spend 20X less on GPUs), then book a call at https://lnkd.in/gMra4-4X

?? ???????????????? ????????????????????????: ???????????? ???? ?????????????? ?????????? ?????????????? ????????????-2! ?? nasscom is thrilled to unveil the selection of 37 ????????????-???????????????? ???????????? ???????????????????? ???? (??????????) ???????????????? for the ???????????? ???????????? ???? ?????? ???????????????????? ???? ?????????????? ??????????????. ?? This marks a significant milestone in India's rapidly evolving #GenAI ecosystem, highlighting the importance of collaborative innovation, improved access to resources, and a strong talent pipeline. ?? Chosen from over 130 applicants, these two-year-old (average age) startups come into the program after having raised $750,000 (on average) in funding. Impressively, 30% of these startups have women founders, reflecting the diversity and inclusivity at the heart of this movement. These startups are creating cutting-edge solutions deployed across various use cases in retail, banking, healthcare, and more. ?? Launched in October 2023, the #NasscomGenAIFoundry program has already supported 26 startups in Cohort-1, forged partnerships with nearly 60 enterprises and investors. These startups have collectively raised $15M. ?? Congratulations to the 37 startups for joining this exciting journey! We look forward to fostering innovation and collaboration in the AI landscape of India with y'all. Ayna, Dresma, XenCapture (XenReality Technologies), Trupeer.ai, immunitoAI, Personaliz.ai, EcoRatings, Sivi, Exora Labs, ulai, Super AI, Catalytics Datum , LEGOAI Technologies, PharynxAI, QuarkGen AI, Waanee AI, Insituate, Xylem AI- Pipeshift (YC S24), Lyzr AI, KushoAI, DaveAI, Fluid AI, Rifa AI, Tecnod8.ai, OnFinance AI, Fundamento, MLAI Digital Private Ltd., Vaani Research Labs, Vitt.AI, Overseas.ai, IntelloSync, KonProz, Boltzmann Labs, Ravity Software Solutions Pvt. Ltd., Pushpak AI, Fruitsalad.ai- GMetriXR, DocuStack #nasscomai #GenerativeAI #Startups #Innovation #WomenInTech ?? Check out the media coverage here: ?? https://lnkd.in/gUKn6NNB ?? https://lnkd.in/gqHrfpRg ?? https://lnkd.in/gE9w_YXz ?? https://lnkd.in/gY5WTnex With Sangeeta Gupta, Ankit Bose

Having trained / finetuned 100s of LLMs for fun and production, I have had my fair bit of struggles ranging from the ever changing training libraries, conflicting package versions to troubles procuring GPUs and burning credits for every step that goes wrong in setting up the training. Always wished for a simple click -> train -> deploy -> that too with a pricing per tokens used structure! Only until the amazing folks Pipeshift AI (YC S24) made it to Y Combinator and launched their much awaited Finetuning and Inference platform for OSS models! I just retrained Kannada Llama and deployed it in an instant with only a few clicks! Thanks to Arko C for giving me access to their beta platform! Do check them out here at https://pipeshift.ai/ #ai #llms #huggingface #training

???Pipeshift AI (YC S24)?launched! Fine-tuning and Inference for Open-Source LLMs "Replace GPT/Claude in production with specialized LLMs that are fine-tuned on your context, offering higher accuracy, lower latencies and model ownership" ?? www.pipeshift.ai ?? Fine-tune Specialized LLMs: Run multiple LoRA-based fine-tuning jobs to build specialized LLMs. ?? Serverless APIs of Base and Fine-tuned LLMs: Run inference for your fine-tuned LLMs and pay as per your token usage. ??? Dedicated Instances for High Speed and Low Latency: Use Pipeshift AI's optimized inference stack to get max throughputs and utilization on GPUs. ?? If you’re building an AI co-pilot/agent/SaaS product and?are looking to move to open-source LLMs?or know someone who’s looking to do the same, reach out via email or?book a call!: https://lnkd.in/gVG2G-cT Congrats Arko C, Enrique Ferrao & Pranav Reddy!! https://fondo.info/3AW5UgI

??????

We presented at Y Combinator Launch Live! Pipeshift’s first ever live demo was in a room of YC founders and the GenAI team at Meta (that too right after a talk by Joseph Spisak). The DevEx came through, the control on deployments came through and most importantly, the vision of making it simple to use and to build with open-source LLMs, came through. Thanks Diana Hu for having us present the demo. The feedback has been amazing since then with multiple companies reaching out to replace GPT/Claude with specialized LLMs that offer higher accuracy, better latency and model control/ownership. PS: If you’re looking to start moving to open-source LLMs, book a call with us here: https://lnkd.in/gMra4-4X Pipeshift AI (YC S24) Pranav Reddy Enrique Ferrao

Let's take Open Source AI to Enterprises!