Konstantinos Krikellas’ Post

There's a super interesting webinar happening this Wednesday. Databricks experts will present a deep dive into Databricks’ new, open source foundation model DBRX. Register for free and learn latest trends in GenAI. #databricks #DBRX #genAI #LLM

FIS has teamed up with Snowflake to bring greater agility and innovation to its capital markets products. With a strong data foundation and better access to insights, FIS is discovering ways to layer AI and ML technologies on top of its data in Snowflake to deliver additional value to customers. Learn how FIS uses 20% less compute and has a 33% cost savings with Snowflake: https://okt.to/Q56t0B

Have personally seen how one of our key product developed sudden interest in our existing customers and prospects and grew significantly after we onboarded our humongous data on snowflake ... one of the best strategic move at the right time ... #FIS #Snowflake #TreasuryAndRisk #CapitalMarkets #InsuranceRiskSuite

With new LLMs being released almost daily, how do you go about navigating such a dynamic landscape? Read below to learn more about the benefits of the agnostic and adaptive nature of Dataiku's "LLM Mesh"

How can financial institutions modernize their core technology at one-tenth of the cost of a traditional legacy migration? For Bendigo and Adelaide Bank, it began by building a global-first use case with generative AI and MongoDB. Watch how MongoDB is critical to their modernization strategy, enabling them to swiftly respond to the needs of their 2.5 million customers: https://lnkd.in/eUDJe-ff

Today is a big day for open-source LLMs! Databricks just unveiled DBRX, the best open source LLM while running 2x faster than Llama-70B. Key features of DBRX: - Efficiency at its core: Its 16x12B MoE architecture means lightning-fast inference and a smaller footprint than similar models. (150 tok/sec) - Top-tier performance: DBRX goes head-to-head with models like GPT-3.5 and Gemini 1.0 Pro in quality, all while being significantly faster. See for yourself! I was blown away by their demo – DBRX handles tough problems like a champ. 👉 https://lnkd.in/gBfF6SZ7 Get the full scoop on the Databricks blog: https://lnkd.in/gzm9hZHr #DBRX #OpenSourceLLM #AI #Databricks

Databricks Mosaic Research team has just released a new general-purpose LLM based on MoE architecture called #DBRX!! #DBRX has scored extremely well in common benchmarks including MMLU, HumanEval, GSM8k, etc. It has beaten out Llama2 and Mixtral variants on many benchmarks showing substantial progress for open source models. I am particularly excited on how this will serve as a base for RAG and fine-tuning including those in both privacy and domain sensitive industries like healthcare & life sciences. Download DBRX today from Hugging Face (DBRX Base, DBRX Instruct), or try out DBRX Instruct in our HF Space, or see our model repository on github: databricks/dbrx. Also, see the announcement on Forbes: https://lnkd.in/eQ5-HWyE #LLM #MoE #Databricks #MosaicAI

👀 Sneak peek into our upcoming TechTalk on AWS Bedrock with Sathya Balakrishnan from AWS and Gabriel C. from Nexla This coming Thursday, learn how to streamline and scale your data integration pipelines for #GenerativeAI with Nexla and #AmazonBedrock. What you’ll learn: - Feed unstructured data into Amazon Bedrock without S3. - Turn text data into vector embeddings and structured data. - Scale data integration for generative AI. - Real-world applications of Nexla’s RAG capabilities. Live demo includes: - Introduction to Nexla and Amazon Bedrock. - Direct data integration methods. - Examples of Nexla’s RAG capabilities enhancing AI. 👉 Register now: https://lnkd.in/ePKdVJUk #TechTalk #AWSBedrock #DataIntegration #GenAI

Get started with your own local AI agent 🤖 Cole Medin will show you how easy it is to get started with a powerful local AI stack that includes Qdrant, alongside Ollama, PostgreSQL, and n8n. He goes beyond the basics, demonstrating live model updates in Docker, Inc (no downtime!), preventing duplicate vectors in Qdrant with custom code, and optimizing RAG performance for standard hardware. 📹 Watch it here: https://lnkd.in/d-NRSn6x 🛠️ See the project on Github: https://lnkd.in/dj7VN-2x

Are you using tools like PyTorch, TensorFlow, Spark, Splunk, Snowflake, Kafka? ... the list goes on 🫶 They all have S3 connectors. Since Cloudian natively speaks S3, we easily integrate with your S3 AI/ML tools to simplify your workflows! Bring the storage team, analytics team, AI team 🤖, research team 🔬, and more together for a centralized data workflow your whole company can agree on! We offer a 🆓 trial cloudian.com/free-trial #ObjectStorage #S3API #S3Storage #S3Data #OnPremObjectStorage #S3ObjectStorage #Kafka #TensorFlow #PyTorch