deepset

Last Chance Webinar! Customized RAG Solutions - Register today 👉🏽 https://hubs.li/Q02W54tn0 Many AI initiatives struggle to get off the ground, taking on average 8 months to move from prototype to deployment. That’s where the partnership between appliedAI Initiative GmbH and deepset steps in. Together, we’re solving the challenges of GenAI for businesses across Europe by driving reliable, Trustworthy RAG solutions that accelerate AI’s impact. Join our joint webinar to learn more: ✅ Real-world approaches to solving AI challenges with Trustworthy RAG ✅ Best practices for implementing AI with transparency and explainability ✅ Lessons from high-stakes industries like electronics and telecom 🚀 Title: Customized RAG Solutions: Rohde & Schwarz’s Success in Knowledge Management for Manufacturing and Telecom 🎙️Speakers: Dr. Christian Geishauser (appliedAI Initiative GmbH), Laura Luckert (deepset) and Andrew Schaefer (Rohde & Schwarz) as they share firsthand insights and actionable strategies from their work with Rohde & Schwarz. 🗓️ Date: 29th of October 2024 ⌚️Time: 4:30pm (1 hour) 📍Place: https://hubs.li/Q02W54tn0 AI’s true value lies in optimizing processes and driving business growth. This event is your chance to learn how Trustworthy RAG can tackle your AI challenges. #Trustworthy #RAG

🚀 (Part IV of our series “Only Custom RAG Is Production-Ready RAG”) Re-ranking makes RAG cheaper, faster, and more accurate Full article 👉 https://lnkd.in/eJMvxpTc 🥇A ranker component helps narrow down the documents we send to the LLM by more accurately ranking the retrieved documents by relevance. 🥈Rankers can use more powerful models than retrievers — or they can use metadata. 🥉In a typical setup, the retriever sends 40-60 documents to the ranker, which re-ranks them and then sends only the top 20 to the LLM. 🏆This saves cost, cuts latency, and reduces noise in the data received by the LLM, improving its performance. Stay tuned for our next post on the most common custom RAG setups! #rag #AI #llm #aiinproduction

Join us for an upcoming webinar based on our latest customer case study! Customized RAG Solutions: Success in Knowledge Management for Manufacturing and Telecom - https://hubs.li/Q02Vq_g80 In collaboration with appliedAI Initiative GmbH, we implemented Retrieval-Augmented Generation (RAG) at Rohde & Schwarz, a leader in electronic products and solutions. Our approach, featuring a robust vision pipeline and retrieval fine-tuning, improved answer accuracy by 46%! The result is a new enhanced, functional prototype that solves challenges such as multilingual data and complex tabular structures and is based on scalable technology for further implementation. This webinar will outline the case study, share learnings, and leave time for questions. 🚀 Title: Customized RAG Solutions: Rohde & Schwarz’s Success in Knowledge Management for Manufacturing and Telecom 🎙️Speakers: Dr. Christian Geishauser at appliedAI, Laura Luckert at deepset, and Andrew Schaefer at Rohde & Schwarz 🗓️ Date: October 29, 2024 ⌚️Time: 4:30pm CET 📍Place: Save your spot here: https://hubs.li/Q02Vq_g80 #GenAI #TrustworthyRAG #RetrievalAugmentedGeneration #customAI

Great read from The Wall Street Journal on how enterprises are moving focus from AI experimentation to delivering tangible business results and ROI. As companies move beyond the initial hype, key challenges are emerging: - "70% of business customers’ generative AI projects are still stuck in the pilot or testing phase" (via Neo4j) - "90% of generative AI experiments don’t make it beyond the lab" due to issues like accuracy and reliability (via Databricks) - Businesses need a "workable prototype that’s fit for purpose and will move the needle" (via Palantir) At deepset, we’re helping companies solve these challenges. deepset Cloud enables fast, accurate prototyping and rapid deployment of AI applications. Our customers, like Manz, YPulse, Siemens, and many more, are building prototypes in days and launching solutions in weeks—delivering real business impact. Check out the blog: https://lnkd.in/ewMXvif6

We’re excited to announce the opening of our new headquarters in Berlin—a milestone made possible by the incredible support of our customers, partners, and community. This growth has enabled us to create a space designed to foster even greater collaboration and innovation, all with the goal of serving you better. Key highlights of the new deepset HQ include: • A central location in Mitte, Berlin’s vibrant heart, where we welcome our Haystack and deepset community, including customers, partners, and team members • 75% of the space is dedicated to collaboration, ensuring we stay close to the needs and ideas of those we serve • A dedicated area for hackathons where we innovate on emerging technologies that matter most to our customers • Hybrid meeting technology designed for inclusive and seamless co-working with global teams and clients • A welcoming environment to build, connect, and inspire new AI solutions Thank you to everyone who has contributed to our journey so far. We look forward to hosting you in our new HQ and continuing to grow together! #HQ #Berlin #innovation #collaboration #inspiredAI

❗ Meet our newest team member, Dominic W., who has joined us in Germany as an Account Executive DACH❗ We asked Dominic, "Why deepset?" 🚀, and here's what he had to say: ⭐ Why deepset: ”I joined deepset driven by a passion for the AI space and its transformative potential. Inspired by deepset’s mission to empower organizations through advanced technology and expertise in natural language processing, I am excited to contribute to a team dedicated to making AI accessible and impactful. With a strong commitment to innovation, I look forward to helping clients unlock new possibilities and drive success through AI-driven insight!” 👤 Bio: Dominic comes with an expertise in sales and technology, particularly within the SaaS environment. Before joining deepset, he worked in the Digital Adoption space progression from account management to senior leadership roles. Bringing a comprehensive understanding of SaaS and AI solutions in the marketplace, he will help our clients and partners to succeed in leveraging the benefit of AI, ultimately contributing to long-term success. ❗ Welcome to deepset, Dominic ❗

Join deepset for this upcoming webinar hosted by partner appliedAI Initiative GmbH and featuring customer Rohde & Schwarz discussing Customized RAG Solutions in Knowledge Management for Manufacturing and Telecom.

🏅 We're proud to be recognized for our innovation in the Gen AI Application development in the annual Accel Euroscape report. Looking forward to the exciting times ahead! 🚀 Thanks for sharing Andrei Brasoveanu https://lnkd.in/g2CKqcsb

💸 GenAI is Transforming Venture Capital Read the complete study: https://lnkd.in/gh3Z4X-Q 🤔 VC’s information finding challenge Our client, a VC firm, had a goldmine of proprietary data, but lacked the tools to effectively use it in day-to-day operations. Traditional enterprise search solutions fell short and they needed a scalable AI strategy. 💡A bespoke solution + fastest time to value Using deepset Cloud, the AI team built a custom LLM-powered tool for the legal department. This AI solution provides instant answers to natural language questions, complete with references to the underlying data. The entire project took less than three months to complete because deepset Cloud allows teams to start from templates and customize as needed. 💪 Instant impact - Tasks that used to take hours are now completed instantly. - Easy access to information from any data source - 60% of team members now use the AI tool as a daily co-pilot - Other departments are requesting similar solutions 📖 Read more The latest customer story is now up on our website: https://lnkd.in/gh3Z4X-Q

❗ Hybrid retrieval is used in the vast majority of RAG pipelines ❗ Full post here: https://lnkd.in/eJMvxpTc (Part III of our series “Only Custom RAG Is Production-Ready RAG”) 💪Different retrievers have different strengths and weaknesses. 🔀Semantic retrievers select documents based on meaning, while keyword-based retrievers identify matching documents based on common words. 🤝To make your retrieval more efficient, you can use the results of different methods to build the context for your LLM. This is called hybrid retrieval. 💡Because of its low effort and high reward, it's become a staple of RAG pipeline setups. Stay tuned for our next post on the most common custom RAG setups! …or read the full article here: https://lnkd.in/eJMvxpTc