Stay up to date with the latest TurboLine releases, deep dives, and product thinking.
Showing 18 of 24 posts
We're launching our WebSocket Examples Directory—a curated collection of 18 production-ready streams from Ethereum, Solana, Polymarket, and major crypto exchanges, complete with connection details and AI analysis prompts.
Introducing TSLN (Time-Series Lean Notation), a text-based serialization format that reduces token usage for time-series data in LLMs by up to 80%, optimizing for how models read numbers.
Comparing latency, token efficiency, and streaming behavior across OpenAI, Anthropic, Azure OpenAI, Gemini, and Grok when processing live BGP update streams.
Our latest pre-production release brings concurrent multi-feed AI analysis, enhanced terminal UI layouts, and critical security hardening for high-throughput data pipelines.
Bring Your Own Model (BYOM) is now supported in TurboStream. Plug in your own LLM provider using simple env vars and keep full control over cost, latency, and model choice.
We're excited to announce Turbostream Terminal UI, our open-source command-line interface for real-time streaming data monitoring and intelligent alerting.
We are excited to announce our pivot to real-time notification system based on streaming data agents, and we're opening up our private beta to select developer partners.
Master Vega and Vega-Lite with essential best practices covering data preparation, chart selection, encoding channels, styling, interactions, performance optimization, and accessibility for publication-quality visualizations.
Using Meta's Segment Anything Model (SAM), we're challenging students to transform sports broadcasting with AI-powered tactical overlays, player positioning analysis, and real-time visual insights.
AI is democratizing movement science. Using computer vision and pose detection, we can now analyze athletic performance, prevent injuries, and provide personalized coaching—all with just a camera.
What if AI could listen to sports and tell the story? Using speech-to-text and language models, we can transcribe match audio and generate enhanced commentary, summaries, and multilingual coverage.
What if AI could capture the magic of sports commentary? Explore how computer vision and language models combine to generate real-time, context-aware sports narration—from detecting plays to storytelling.
TurboLine.ai partners with IIMS College to host an international hackathon, bridging academia and industry. This collaboration empowers Nepalese students with real-world tech challenges and global networking opportunities.
Turboline is now a Microsoft ISV Partner! Our Data Studio and AI Data Engine™ are available on Azure Cloud Marketplace, bringing enterprise-ready solutions directly to your Azure infrastructure.
We're proud to partner with Silicon Technix, a London-based digital agency specializing in web design, development, and AI consultancy since 2008. Together, we're driving innovation and growth.
18 applications, 10 innovative teams, and 2 industry judges came together for our AI and Data hackathon. D Blunt won with their AI-powered travel planning app after an exciting tie-breaker.
Discover data-driven strategies to boost your NuGet package visibility. Our analysis of 59 packages reveals that tags and keyword optimization matter far more than downloads for ranking higher in search results.
A detailed comparison of Mistral Large and Google Gemini 1.5 Pro's capabilities in generating PySpark code for data transformation tasks, analyzing code quality, efficiency, and practical implementation.
