🚀 𝐑𝐞𝐚𝐥-𝐓𝐢𝐦𝐞 𝐀𝐈 𝐈𝐬 𝐇𝐞𝐫𝐞 𝐭𝐨 𝐒𝐭𝐚𝐲 🚀 2025 is here, and if there’s one thing we learned in 2024, it’s this: real-time AI isn’t optional anymore. Last year, we wrote about 𝘙𝘦𝘵𝘳𝘪𝘦𝘷𝘢𝘭 𝘈𝘶𝘨𝘮𝘦𝘯𝘵𝘦𝘥 𝘎𝘦𝘯𝘦𝘳𝘢𝘵𝘪𝘰𝘯 (𝘙𝘈𝘎) and how it’s reshaping AI systems. Even now, we believe our blog remains incredibly relevant. RAG is about solving a problem we’ve all faced: How do you make sure your AI has access to the most 𝘶𝘱-𝘵𝘰-𝘥𝘢𝘵𝘦 𝘪𝘯𝘧𝘰𝘳𝘮𝘢𝘵𝘪𝘰𝘯? The answer: ✔ You retrieve the latest data from a live database. ✔ You let AI generate answers that are accurate and grounded in that data. ✔ And you make it all work in real-time with tools like Bytewax. 🐝 This blog digs into the details of building scalable, real-time AI workflows. If you haven’t read it yet, it’s worth a look: 🔗 Here’s the link: bytewax.io/blog/building-rea… P.S. What’s one thing you think AI still struggles with in real-time applications

🚀 Transforming Bytewax with LLM Power What’s the difference between an agent and a chatbot? Action. @LGFunderburk latest blog shows you how to equip Bytewax dataflows to work alongside LLMs. “Agents don’t just reply—they decide, plan, and execute.” ✅ Automate decision-making ✅ Build smarter workflows ✅ Use frameworks like @langchain and @deepset_ai The full blog is here: bytewax.io/blog/using-llms-w…

🌟 2024 in Review: Built by Bytewax, Powered by Community 🌟 The third chapter of our Year in Review is here! This time, we’re focusing on what’s been built with Bytewax, the tools we’ve created, and the conversations we’ve had. 💡 Built with Bytewax - The "Used By" feature on GitHub reveals amazing projects powered by Bytewax this year, like: 🌲 Air Quality Anomaly Detection with @redpandadata. 📰 Real-Time News Search combining Bytewax, @apachekafka, @upstash, and @langchain. 💹 Finance Pipelines using Bytewax with @hopsworks to process live trade data. ✨ Shoutout to Spanda AI Inc. for human-centered AI solutions and real-time embeddings! 📚 Bytewax Cheatsheets - We launched three cheatsheets this year: 1️⃣ 𝐁𝐲𝐭𝐞𝐰𝐚𝐱 𝐁𝐚𝐬𝐢𝐜𝐬: Master dataflows, connectors, and Python-powered transformations. 2️⃣ 𝐎𝐩𝐞𝐫𝐚𝐭𝐨𝐫𝐬: From map to stateful_batch, dive into real-world examples. 3️⃣ 𝐖𝐢𝐧𝐝𝐨𝐰𝐢𝐧𝐠: Tumbling, sliding, session windows—streaming made manageable. 🎙️ Podcasts and Events 2024 wasn’t just about tech—it was about people: 💡 OSS4AI Meetup: Laura Funderburk shared Bytewax’s approach to RAG, supported by @olidinov during her maternity leave. 🌍 Softlandia Ltd. Workshop: 480+ registrants, 10 countries, deep dives into real-time RAG pipelines. 🎙️ @Redisinc Podcast: @MathesonZander discussed Bytewax’s developer-first approach. 🚀 This year was about building, learning, and collaborating—and it’s all thanks to YOU. 💛 Read more here: bytewax.io/blog/bytewax-2024…

🛠️ 𝐍𝐞𝐰 𝐂𝐨𝐧𝐧𝐞𝐜𝐭𝐨𝐫𝐬 𝐢𝐧 𝐭𝐡𝐞 𝐁𝐲𝐭𝐞𝐰𝐚𝐱 𝐌𝐨𝐝𝐮𝐥𝐞 𝐇𝐮𝐛: 𝐁𝐮𝐢𝐥𝐭 𝐛𝐲 𝐎𝐮𝐫 𝐂𝐨𝐦𝐦𝐮𝐧𝐢𝐭𝐲 Our stream processor’s flexibility allows developers to build custom sources and sinks tailored to their needs. This adaptability has inspired our community to create tools that benefit everyone. 𝘛𝘩𝘦 𝘭𝘢𝘵𝘦𝘴𝘵 𝘢𝘥𝘥𝘪𝘵𝘪𝘰𝘯𝘴 𝘵𝘰 𝘵𝘩𝘦 𝘔𝘰𝘥𝘶𝘭𝘦 𝘏𝘶𝘣 𝘪𝘯𝘤𝘭𝘶𝘥𝘦: • @webflow: A sink connector for inserting or updating records in a specified Webflow Collection. • @valkey_io: Source and sink connectors for Valkey, an open-source in-memory datastore. • Lever: A connector that pulls new job postings from the Lever job platform. 💛 We’re proud to see our ecosystem grow through the ingenuity of developers who contribute back to the community. 𝐇𝐚𝐯𝐞 𝐚 𝐜𝐨𝐧𝐧𝐞𝐜𝐭𝐨𝐫 𝐭𝐨 𝐬𝐡𝐚𝐫𝐞? Let us help you bring it to the broader developer world! (𝘓𝘦𝘢𝘷𝘦 𝘢 𝘤𝘰𝘮𝘮𝘦𝘯𝘵, 𝘰𝘳 𝘫𝘶𝘴𝘵 𝘋𝘔 𝘶𝘴) ➡️ bytewax.io/modules

🚛 This isn’t your average trucking story. @RangeEnergy is electrifying trailers, but that’s only half the tale. With Bytewax, they’ve built real-time pipelines that turn GPS data into live experiments. Forget slow-moving legacy systems—this team’s rewriting the rules: ➡️ Python-first solutions, because why overcomplicate things? ➡️ Real-time telematics, from road to cloud in seconds. ➡️ Scalable dataflows that grow with their ambition. 📖 Curious? Jump in here: bytewax.io/blog/bytewax-rang…

𝙁𝙤𝙧𝙜𝙚𝙩 𝙤𝙪𝙩𝙙𝙖𝙩𝙚𝙙 𝙗𝙖𝙩𝙘𝙝 𝙥𝙧𝙤𝙘𝙚𝙨𝙨𝙞𝙣𝙜—𝙮𝙤𝙪𝙧 𝙍𝘼𝙂 𝙬𝙤𝙧𝙠𝙛𝙡𝙤𝙬𝙨 𝙙𝙚𝙨𝙚𝙧𝙫𝙚 𝙗𝙚𝙩𝙩𝙚𝙧. 🧐 Laura Funderburk, our expert on all things RAG, has written a detailed step-by-step blog on creating real-time APIs using Bytewax, @Haystack_AI , and @FastAPI . In this post, Laura explains how we: ✔️ Indexed data with Bytewax to handle dynamic, large-scale datasets efficiently. ✔️ Used Haystack to implement embedding-based retrieval for fast and accurate results. ✔️ Leveraged FastAPI to design responsive, queryable endpoints that ensure low latency under heavy workloads. The blog also covers challenges like efficiently indexing large datasets, ensuring low-latency retrieval, and scaling to meet variable workloads. Through a real-world case study, you’ll see how these tools come together in a framework that delivers relevance and performance in real time. 📖 Full blog: bytewax.io/blog/rag-app-case…

𝘓𝘦𝘵’𝘴 𝘊𝘰𝘯𝘵𝘪𝘯𝘶𝘦 𝘖𝘶𝘳 2024 𝘙𝘦𝘧𝘭𝘦𝘤𝘵𝘪𝘰𝘯𝘴!🚀 This year wasn’t just about innovation—it was about 𝙘𝙤𝙢𝙢𝙪𝙣𝙞𝙩𝙮, 𝙘𝙤𝙡𝙡𝙖𝙗𝙤𝙧𝙖𝙩𝙞𝙤𝙣, and building 𝙘𝙤𝙣𝙣𝙚𝙘𝙩𝙞𝙤𝙣𝙨 💛. 🎉 GenAI & Bytewax: Real-Time Meets Real Innovation From bytewax-@Redis to real-time RAG with Microsoft Azure , we’ve made huge strides in making real-time workflows scalable and efficient. Special thanks to @iusztinpaul, Alex Razvant, and Alex Vesa for their contributions to the LLM Twin Course, which has empowered so many developers to build production-ready AI replicas. 🌍 Global Connections at @PyData We were honored to join @NumFOCUS as a startup partner this year, participating in PyData London and PyData NYC, where we connected with brilliant minds and showcased Bytewax’s unique strengths as a Python-native stream processor. 🍕 Real-Time Pizza Analytics Workshop Collaborating with @startreedata @gAmUssA and @streamlit @CarolineFrasca, we explored real-time analytics using Bytewax, Apache Pinot, and Streamlit to build a pizza dashboard. Real-time data has never been more fun (or delicious). 🐝 Swag and Stories: From Chonky Bee keychains to lunch boxes, Bytewax swag became a global symbol of our connection with you. 📖 Read the Full Blog Here: bytewax.io/blog/real-time-ai…

🚨 𝐍𝐞𝐰 𝐁𝐲𝐭𝐞𝐭𝐚𝐥𝐤𝐬 𝐄𝐩𝐢𝐬𝐨𝐝𝐞🚨 ☝️ Yesterday, we shared the big news: Bytewax now integrates with @duckdb and @motherduck , bridging the worlds of real-time stream processing and high-performance analytics. To give you all the details, we recorded a special episode of Bytetalks 🐝 with @matsonj, a Developer Advocate from MotherDuck. 🦆 🎙 Here’s what you’ll learn in this episode: ✔ The role of the Bytewax DuckDB/MotherDuck connector in transforming data workflows. ✔ How to seamlessly handle both real-time and batch data scenarios. ✔ Use cases that showcase the power of local-first analytics and cloud collaboration. 📖 This integration is a solution for the growing demand for flexible, efficient data pipelines.  As Jacob shares, “𝘞𝘪𝘵𝘩 𝘔𝘰𝘵𝘩𝘦𝘳𝘋𝘶𝘤𝘬 𝘢𝘯𝘥 𝘉𝘺𝘵𝘦𝘸𝘢𝘹, 𝘺𝘰𝘶 𝘤𝘢𝘯 𝘸𝘰𝘳𝘬 𝘧𝘢𝘴𝘵𝘦𝘳 𝘢𝘯𝘥 𝘴𝘮𝘢𝘳𝘵𝘦𝘳—𝘯𝘰 𝘮𝘢𝘵𝘵𝘦𝘳 𝘪𝘧 𝘺𝘰𝘶’𝘳𝘦 𝘳𝘶𝘯𝘯𝘪𝘯𝘨 𝘭𝘰𝘤𝘢𝘭𝘭𝘺 𝘰𝘳 𝘴𝘤𝘢𝘭𝘪𝘯𝘨 𝘪𝘯 𝘵𝘩𝘦 𝘤𝘭𝘰𝘶𝘥" Watch the episode here: youtu.be/QK2GVAA2Zy0?si=MvSC…

🔥 𝐁𝐫𝐢𝐧𝐠𝐢𝐧𝐠 𝐒𝐭𝐫𝐞𝐚𝐦 𝐏𝐫𝐨𝐜𝐞𝐬𝐬𝐢𝐧𝐠 𝐚𝐧𝐝 𝐀𝐧𝐚𝐥𝐲𝐭𝐢𝐜𝐬 𝐂𝐥𝐨𝐬𝐞𝐫 𝐓𝐨𝐠𝐞𝐭𝐡𝐞𝐫 Real-time data demands real-time solutions. That’s why we’re excited to announce Bytewax’s integration with @duckdb and @motherduck, combining the power of stream processing with high-performance analytics. This integration offers practical new capabilities: 📊 𝐑𝐞𝐚𝐥-𝐓𝐢𝐦𝐞 𝐃𝐚𝐬𝐡𝐛𝐨𝐚𝐫𝐝𝐬: Seamlessly stream metrics into DuckDB for live insights. 🌐 𝐇𝐲𝐛𝐫𝐢𝐝 𝐖𝐨𝐫𝐤𝐟𝐥𝐨𝐰𝐬: Merge on-premises processing with cloud scalability using MotherDuck. ⚙️ 𝐄𝐟𝐟𝐢𝐜𝐢𝐞𝐧𝐭 𝐄𝐓𝐋 𝐏𝐢𝐩𝐞𝐥𝐢𝐧𝐞𝐬: Stream, transform, and store data in real-time without complexity. 𝘛𝘩𝘪𝘴 𝘪𝘯𝘵𝘦𝘨𝘳𝘢𝘵𝘪𝘰𝘯 𝘪𝘴 𝘢 𝘱𝘳𝘢𝘤𝘵𝘪𝘤𝘢𝘭 𝘴𝘰𝘭𝘶𝘵𝘪𝘰𝘯 𝘧𝘰𝘳 𝘣𝘶𝘪𝘭𝘥𝘪𝘯𝘨 𝘥𝘢𝘴𝘩𝘣𝘰𝘢𝘳𝘥𝘴 𝘢𝘯𝘥 𝘳𝘦𝘧𝘪𝘯𝘪𝘯𝘨 𝘌𝘛𝘓 𝘸𝘰𝘳𝘬𝘧𝘭𝘰𝘸𝘴. Find out how it works in our own @LGFunderburk's latest blog ➡️ bytewax.io/blog/bytewax-duck…

"𝘈𝘐 𝘸𝘪𝘭𝘭 𝘯𝘦𝘷𝘦𝘳 𝘳𝘦𝘱𝘭𝘢𝘤𝘦 𝘵𝘩𝘦 𝘬𝘯𝘰𝘸𝘭𝘦𝘥𝘨𝘦 𝘰𝘧 𝘣𝘶𝘪𝘭𝘥𝘪𝘯𝘨 𝘳𝘰𝘣𝘶𝘴𝘵, 𝘴𝘤𝘢𝘭𝘢𝘣𝘭𝘦, 𝘢𝘯𝘥 𝘳𝘦𝘭𝘪𝘢𝘣𝘭𝘦 𝘴𝘺𝘴𝘵𝘦𝘮𝘴." Our very own @LGFunderburk recently joined @thechrislusk from @AIMakerspace to share her journey, insights on Gen AI, and the challenges of real-time RAG systems. 🚀 She spoke about her passion for integrating generative AI into real-time applications and how it’s reshaping what’s possible with streaming analytics. 🤫 Laura also offered an exciting glimpse of her upcoming book, where she demystifies building robust RAG systems with Bytewax and Haystack (@deepset_ai). The book focuses on practical, real-world use cases backed by her extensive experience with LLMs and streaming technologies. Catch the full interview and stay updated on her groundbreaking work! youtube.com/watch?v=rtoBqacZ… 💛 Bytewax is proud to have Laura on the team, pushing the boundaries of what’s possible in real-time AI.

🐝 Bytetalks is officially here, and we’re excited to present our next episode! So, today, our own @LGFunderburk and @MathesonZander will take you through Bytewax operators, the key components driving streaming dataflows. Watch to learn: ☒ The basics of Bytewax dataflows, including workers, clusters, and directed graphs. ☒ How stateless and stateful operators work, and why managing the state is crucial. ☒ Practical examples like filtering smoothie orders, enriching data with pricing, and calculating totals with tax. ☒ Insights on keying data for stateful processing and how caching can improve performance. Join us as we explore all things streaming! The full episode ➡️ youtube.com/watch?v=F_oHB1r7…

📣 New Tool, New Possibilities! 📣 The 𝐁𝐲𝐭𝐞𝐰𝐚𝐱 𝐈𝐧𝐟𝐥𝐮𝐱𝐃𝐁 𝐂𝐨𝐧𝐧𝐞𝐜𝐭𝐨𝐫 is here, making it easier than ever to connect, process, and visualize your time-series data in real-time. 😎 No more waiting around for delayed insights — get your data flowing smoothly between Bytewax and InfluxDB for predictive analytics, anomaly detection, and more. Check out our blog post for a detailed guide on getting started with the Bytewax InfluxDB Connector, complete with setup tips, code snippets, and practical use cases. 👀 Learn how to easily write data to @InfluxDB using Bytewax’s InfluxDBSink and pull real-time streams directly with InfluxDBSource. The full article ➡️ bytewax.io/blog/bytewax-infl…