Tavily

Join Guy Hartstein and I to build the client-facing eng team from the ground up. dm if interested

Tavily is now available on AWS Marketplace ?? Enterprise-grade web infrastructure to power your AI agents. Search, Extract, Crawl - all in one. Secure. Reliable. At Scale. ??? https://lnkd.in/dcHEPM7h Built for AI-native teams. Ready for production. Amazon Web Services (AWS)

Excited to share that Tavily is part of NVIDIA’s new AI-Q Blueprint, announced at #GTC2025! Our retrieval tech is powering the next generation of agentic systems — enabling AI agents to reason, collaborate, and move fast across web and enterprise data. Huge thanks to the NVIDIA team for including us in this milestone. The future of work is agentic — and it’s here. ?? https://lnkd.in/d9VsjbYk

Cline + Tavily = AI Code Assistant, Supercharged Add the Tavily MCP from the Cline Marketplace for real-time answers from GitHub issues, docs, and Stack Overflow—delivered instantly into VS Code with one-click installation. Stay focused with ready-to-implement code, faster debugging, and seamless AI assistance—no more context switching ?? ?? Full guide by Dustin Pulver https://lnkd.in/datPVgby ?? Tavily MCP repo:?https://lnkd.in/dw4fq8yC

search. extract. crawl. coming soon join the waitlist https://crawl.tavily.com/

Recursive Retrieval & Reasoning: Enhancing RAG with Reasoning??? Traditional RAG is powerful, but retrieving the right context isn’t just about fetching documents—it requires reasoning to refine queries, assess relevance, and iteratively improve answers. Recursive Retrieval with Reasoning (RRR)?by?Dean Sacoransky & Deniz Askin, Ph.D.?improves retrieval by applying a reasoning model at every step to: ?- Identify gaps in retrieved info ?- Dynamically refine queries ?- Ensure only relevant sources are used With recursive reasoning, you can iteratively search for information to tackle complex questions—like?“Is RBC’s headquarters north of Sam Altman’s brother’s company headquarters?” See how DeepSeek’s reasoning traces determine which sources matter, which don’t, and what additional queries are needed in the blog below ?? ?? https://lnkd.in/d6XiwFWu LangChain DeepSeek AI

?? Excited to share our latest open-source project. Lorenzo Micheli and I have just released an aws-samples repository for Bedrock Deep Researcher: https://lnkd.in/dMRBcTVj - an Agentic application that helps you generate high-quality articles and blog posts, complete with custom banner images. ? Key features: - Powered by Anthropic Claude 3.5 Haiku, and 3.7 Sonnet (or you can choose any other model of your choice from Amazon Bedrock) for planning, query and content generation. - Web search integration using Tavily API for real-time research. - Streamlined article generation with custom instructions. - Automatic banner image creation for your content using Amazon Nova Canvas models. - Easy-to-use #Streamlit interface This open-source tool built using #LangGraph and #AmazonBedrock is perfect for content creation, technical documentation, blog posts and for researchers looking to streamline their content generation workflow while maintaining quality and accuracy. Check it out??? ?? We would love your feedback and contributions! ?? #AmazonBedrock #Anthropic #AgenticAI #ContentCreation #OpenSource #GenerativeAI

This is one of the best (open-source) research agents l've seen in a while, especially Ul-wise. Made by Guy Hartstein, this company research agent uses grounding and document curation for maximum accuracy. It utilizes multiple models and parallelizes processes to maximize efficiency, streaming the process and final report to an engaging frontend. A lot of great techniques here with LangGraph-based agents in the repo. Really recommend checking it out:?https://lnkd.in/dJHc_GhZ LangChain Tavily

???? Tavily + Cats with Bats: Keeping Developers Focused Developers learn best when they stay in the flow—no distractions and endless tab-switching. Cats with Bats, an AI-powered platform for developer education, integrates Tavily Web Search to supercharge its AI tutor ?? When learners ask questions beyond a lesson’s scope, the AI tutor first searches the course. If no answer is found, Tavily delivers accurate, real-time responses directly within the platform. Stay tuned for Tavily Academy coming soon on www.catswithbats.com Hai N.

?? Introducing Deep Research for GPT Researcher! We're excited to announce our latest feature that takes autonomous research to the next level! Deep Research is an advanced recursive system that explores topics with unprecedented depth and breadth. ?? What makes it special: ? Tree-like exploration with configurable depth and breadth ? Concurrent processing for lightning-fast results ? Smart context management across research branches ? Real-time progress tracking ? Comprehensive synthesis of findings ?? Performance that matters: ? Complete deep research in ~5 minutes ? Cost-effective at ~$0.4 per research ? Powered by o3-mini with high reasoning effort ?? How it works: Think of it as deploying a team of specialized AI researchers, each diving deep into different aspects of your topic while maintaining a cohesive understanding of the bigger picture. The system automatically aggregates and synthesizes findings across all branches, ensuring no valuable insight is missed. ?? Getting Started: 1. Install via pip: gpt-researcher 2. Select "Deep Research" mode 3. Watch as it explores your topic in ways never before possible To learn more: ?? Documentation: https://lnkd.in/d9mTp8gN ?? Join our community: https://lnkd.in/dQv_3mwY