deepset_ai Profile Banner
deepset Profile
deepset

@deepset_ai

Followers
5K
Following
2K
Media
194
Statuses
1K

We provide open source technology and products to interact with your data using LLMs and build production-ready AI Applications. Creators of @Haystack_AI

Berlin, Germany
Joined December 2017
Don't wanna be here? Send us removal request.
@deepset_ai
deepset
8 months
Excited to share that deepset has been recognized as a 2024 Gartnerยฎ Cool Vendor in AI Engineering. We believe this recognition validates our innovative AI approach and work empowering enterprise AI adoption. Get the report:
Tweet media one
0
2
14
@deepset_ai
deepset
7 days
RT @Haystack_AI: ๐Ÿ”ฅ Haystack 2.15 just landed with faster agents and AI guardrails for safer apps. Hereโ€™s whatโ€™s new ๐Ÿ‘‡. โšก Parallel Tool Callโ€ฆ.
0
2
0
@deepset_ai
deepset
19 days
RT @Haystack_AI: ๐Ÿš€ You can use Gemini 2.5 Pro, Flash, and Flash-Lite with Haystack! ๐Ÿ”ฅ. Build smart, flexible pipelines and agents with yourโ€ฆ.
0
5
0
@deepset_ai
deepset
27 days
Today at #GTCParis weโ€™reย announcing our integration with @nvidia Enterprise AI Factory validated design, enabling production-ready AI systems tailored to enterprise requirements, delivered on-premises. Organizations can now deploy sophisticated multi-agent workflows on NVIDIA
Tweet media one
0
0
1
@deepset_ai
deepset
1 month
๐Ÿ“ข Excited to share our new case study with @awscloud and @AnthropicAI featuring MANZ. MANZ Genjus KI is an AI-powered legal assistant that transforms how legal professionals research, analyze documents, and draft contracts. This innovative AI agent:. โ€ข Thoroughly analyzes
Tweet media one
0
0
1
@deepset_ai
deepset
1 month
RT @Haystack_AI: ๐Ÿ”” ๐—›๐—ฎ๐˜†๐˜€๐˜๐—ฎ๐—ฐ๐—ธ ๐Ÿฎ.๐Ÿญ๐Ÿฐ ๐—ถ๐˜€ ๐—ต๐—ฒ๐—ฟ๐—ฒ โ€” ๐˜„๐—ถ๐˜๐—ต ๐—ฟ๐—ฒ๐—ฎ๐—น-๐˜๐—ถ๐—บ๐—ฒ ๐—ฎ๐—ด๐—ฒ๐—ป๐˜ ๐˜€๐˜๐—ฟ๐—ฒ๐—ฎ๐—บ๐—ถ๐—ป๐—ด ๐—ฎ๐—ป๐—ฑ ๐—ฑ๐—ฒ๐—ฒ๐—ฝ๐—ฒ๐—ฟ ๐—ฝ๐—ถ๐—ฝ๐—ฒ๐—น๐—ถ๐—ป๐—ฒ ๐˜ƒ๐—ถ๐˜€๐˜‚๐—ฎ๐—น๐—ถ๐˜‡๐—ฎ๐˜๐—ถ๐—ผ๐—ป๐˜€. This release brings powerfulโ€ฆ.
0
2
0
@deepset_ai
deepset
2 months
RT @Haystack_AI: ๐Ÿ–ผ๏ธ Building agents that understand both text and images?. We've released @Haystack_AI's new experimental Multimodal supporโ€ฆ.
0
3
0
@deepset_ai
deepset
2 months
RT @Haystack_AI: ๐Ÿš€ Deploying AI pipelines just got way easier!. ๐—›๐—ฎ๐˜†๐˜€๐˜๐—ฎ๐—ฐ๐—ธ makes it easy to build modular LLM pipelines for agentic apps. Nowโ€ฆ.
0
2
0
@deepset_ai
deepset
2 months
๐Ÿ› ๏ธ RAG may not be the most hyped AI architecture right now. But it remains the one that delivers the most value to users. Gen AI evolves fast, but RAG remains one of the most reliable, scalable, and industry-ready approaches for bringing LLMs into the real world. Whether you're
Tweet media one
0
4
10
@deepset_ai
deepset
2 months
๐Ÿ“š NEW RESOURCE: ๐™๐™š๐™ฉ๐™ง๐™ž๐™š๐™ซ๐™–๐™ก-๐˜ผ๐™ช๐™œ๐™ข๐™š๐™ฃ๐™ฉ๐™š๐™™ ๐™‚๐™š๐™ฃ๐™š๐™ง๐™–๐™ฉ๐™ž๐™ค๐™ฃ ๐™ž๐™ฃ ๐™‹๐™ง๐™ค๐™™๐™ช๐™˜๐™ฉ๐™ž๐™ค๐™ฃ ๐™ฌ๐™ž๐™ฉ๐™ ๐™ƒ๐™–๐™ฎ๐™จ๐™ฉ๐™–๐™˜๐™  by O'Reilly. Retrieval-augmented generation (RAG) has become the de facto standard for ensuring the factual accuracy of generative AI models.
1
0
3
@deepset_ai
deepset
2 months
Weโ€™ve partnered with @weights_biases Weave to bring enterprise-grade observability to AI apps and agents built on the deepset AI Platform. This collaboration combines the best of both worlds: deepsetโ€™s flexible, modular framework for building AI applications, and Weights &.
0
0
5
@deepset_ai
deepset
3 months
๐Ÿ”” @Haystack_AI ๐Ÿฎ.๐Ÿญ๐Ÿฏ.๐Ÿฌ ๐—ถ๐˜€ ๐—ต๐—ฒ๐—ฟ๐—ฒ โ€” ๐˜„๐—ถ๐˜๐—ต ๐—ฎ๐—ด๐—ฒ๐—ป๐˜ ๐˜๐—ฟ๐—ฎ๐—ฐ๐—ถ๐—ป๐—ด ๐—ฎ๐—ป๐—ฑ ๐—ฒ๐—ป๐—ต๐—ฎ๐—ป๐—ฐ๐—ฒ๐—ฑ ๐˜๐—ผ๐—ผ๐—น ๐—บ๐—ฎ๐—ป๐—ฎ๐—ด๐—ฒ๐—บ๐—ฒ๐—ป๐˜!. Hereโ€™s whatโ€™s new:. ๐Ÿ” ๐—”๐—ด๐—ฒ๐—ป๐˜ ๐—ง๐—ฟ๐—ฎ๐—ฐ๐—ถ๐—ป๐—ด.Get deeper visibility into your agentโ€™s behavior. Each step in the Agent is now traced and
Tweet media one
0
1
4
@deepset_ai
deepset
3 months
RT @Haystack_AI: ๐Ÿ”” ๐—›๐—ฎ๐˜†๐˜€๐˜๐—ฎ๐—ฐ๐—ธ ๐Ÿฎ.๐Ÿญ๐Ÿฏ.๐Ÿฌ ๐—ถ๐˜€ ๐—ต๐—ฒ๐—ฟ๐—ฒ โ€” with agent tracing and enhanced tool management!. With powerful agent tracing, smarter tool maโ€ฆ.
0
1
0
@deepset_ai
deepset
3 months
๐Ÿšจย New White Paper: Transform Complex Document Analysis Workflows with Generative #AI. Intelligent Document Processing (#IDP) is top of mind for many of our customers, and for a simple reason: organizations' document collections are constantly growing and need to be leveraged to
Tweet media one
0
1
3
@deepset_ai
deepset
3 months
RT @Haystack_AI: ๐—ฆ๐—ฝ๐—ฟ๐—ถ๐—ป๐—ด ๐—ถ๐—ป๐˜๐—ผ ๐—›๐—ฎ๐˜†๐˜€๐˜๐—ฎ๐—ฐ๐—ธ: ๐—ข๐˜‚๐—ฟ ๐—ฆ๐—ฒ๐—ฎ๐˜€๐—ผ๐—ป๐—ฎ๐—น ๐—–๐—ต๐—ฎ๐—น๐—น๐—ฒ๐—ป๐—ด๐—ฒ ๐—ถ๐˜€ ๐—Ÿ๐—ถ๐˜ƒ๐—ฒ! ๐ŸŒธ. Looking for a reason to ๐˜ฅ๐˜ช๐˜จ ๐˜ช๐˜ฏ๐˜ต๐˜ฐ ๐˜ˆ๐˜ ๐˜ˆ๐˜จ๐˜ฆ๐˜ฏ๐˜ต๐˜ด and explore the ๐— ๐—ผ๐—ฑ๐—ฒ๐—นโ€ฆ.
0
1
0
@deepset_ai
deepset
3 months
There's a new standard in town ๐Ÿ”ฅ. Model Context Protocol (MCP) promises to change how we build AI applications by standardizing access to tools and data. But what makes it actually useful? Our new blog breaks down #MCP with real examples that show how it works AND why it
Tweet media one
0
0
3
@deepset_ai
deepset
3 months
RT @Haystack_AI: ๐Ÿšจ Haystack Weekly Updates ๐Ÿšจ. ๐Ÿ“ข Haystack 2.12 is here โ€” meet the new Agent component!.Build tool-calling agents, reuse fullโ€ฆ.
0
1
0
@deepset_ai
deepset
3 months
Weโ€™re excited to share our latest case study with @AirbusDefence and ๐Ÿš€on the KITCH studyโ€”a Sovereign AI solution developed for the German Armed Forces (@bundeswehrInfo). This system uses #AI to safely analyze military regulations and run combat simulations, helping leaders
Tweet media one
0
0
3
@deepset_ai
deepset
3 months
We had an incredible day at @DeepLearningAI 's #AIDev25 last Friday! Our team was energized by all the brilliant conversations at our booth with developers building cutting-edge AI applications. Huge thanks to @AndrewYNg and the entire team for creating this space for the AI.
@AndrewYNg
Andrew Ng
4 months
Last Friday on Pi Day, we held AI Dev 25, a new conference for AI Developers. Tickets had (unfortunately) sold out shortly after we announced their availability, but I came away energized by the day of coding and technical discussions with fellow AI Builders! Let me share here my
0
0
0
@deepset_ai
deepset
3 months
๐Ÿ’กย Generative AI can reduce the time spent on high-volume financial knowledge work by 40%, while maintaining an astounding 99.5% correctness. Financial services platform credX transformed its lending process by expediting the analysis of over 1,000 documents with Gen AI, freeing
Tweet media one
0
0
2