Raziel Tabib

Today at Rumi Labs, Inc. we launched our Desktop App, running our proprietary SOTA edge VLM optimized for understanding complex multi-modal narratives. This makes Rumi the 1st-ever distributed Vision-Language Model network for Media in production. If you want to help build the Narrative Intelligence Layer for AI, download the Rumi extension at http://rumi.io

13

2 Comments

Hot take: 𝗡𝗼 𝗼𝗻𝗲'𝘀 𝗯𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗺𝘂𝗹𝘁𝗶-𝗺𝗼𝗱𝗮𝗹 𝗮𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗰𝗼𝗿𝗿𝗲𝗰𝘁𝗹𝘆. WWDC and Google I/O introduced new paradigms, but apps are still lagging behind general best practices. Here are 4️⃣ ways to catch up and embrace what's new. 1️⃣ You'll never build a successful app without embracing platform-specific design systems. Every app we've built that ignored these conventions came with a ton of friction:  • Custom component and UI maintenance  • Inconsistent user experiences  • Constant updates to "catch up" to platform evolution If your app isn't gearing up for Liquid Glass or Material 3 Expressive, there's zero chance you're optimizing for success. 2️⃣ On-device AI is going to be a huge product differentiator. I thought I understood on-device LLM integration, but seeing the new capabilities gives me a fresh perspective on privacy-first, low-latency, highly personalized experiences. Treat on-device AI as a core product feature: study the APIs, experiment with how models can make your products better, and make it central to your app's value proposition. Composable architecture and UI has a really bright future when it's augmented with AI. 3️⃣ Systems integration is the most underrated engineering challenge. We've built hundreds of apps, and I can probably count on one hand where we truly leveraged deep system integration. We spend hours on custom features while ignoring things like App Intents, Live Activities, Focus Filters, etc. These are all areas that could drive higher user engagement and satisfaction. Surprise and delight, always; if I were a product manager, I'd challenge the roadmap to replace a custom feature with an existing platform capability. 4️⃣ True multi-device support is the future. It's clear that each device and form factor is its own platform, and responsive design is no longer enough. Each device needs tailored experiences. It's not about scaling UI, it's about rethinking functionality for each form factor. Sound overwhelming? It can be...𝘣𝘶𝘵 𝘪𝘵'𝘴 𝘢𝘭𝘴𝘰 𝘵𝘩𝘦 𝘮𝘰𝘴𝘵 𝘦𝘹𝘤𝘪𝘵𝘪𝘯𝘨 𝘵𝘪𝘮𝘦 𝘵𝘰 𝘣𝘦 𝘣𝘶𝘪𝘭𝘥𝘪𝘯𝘨 𝘢𝘱𝘱𝘴! Want to dive deeper into these insights? Check out the full article I co-authored: https://lnkd.in/gARcXHA9 Or better yet, let's connect. I'm always eager to exchange ideas with fellow tech leaders navigating this new landscape.

35

1 Comment

Always fun to host engineering dinners with Sumeet Vaidya and Crafting. Few takeaways for me: 1. We are living through one of the most dynamic periods in software engineering. Everyone by now knows to heavily leverage “Plan” mode upfront when coding with AI - but now we are hearing there’s even more time/tokens/turns spent on “Ask” mode before writing the Plan. And PRs of plans! We are literally flying up the abstractions curve. 2. Documentation used to be a luxury, but it is now cheap and easy. Documentation is increasingly agent-first, and markdowns in the repo seem to work a lot better than Confluence/Notion. 3. Historically the magic of GitHub was how it captured the collaboration context - PRs, discussions, reviews, commit history. Today so much of that context sits in prompt/chat history which does not go in GitHub or anywhere else! Feels like this needs to be solved. Openclaw’s memory system and Granola are both inspirations.

13

1 Comment

Super excited about the pace at which Generative AI is reshaping video and multilingual storytelling for newsrooms! With platforms like OpenAI Sora, Runway Gen-2, HeyGen, ElevenLabs, and Pika, creating localized, publisher-ready video content is now faster, more natural, and more scalable than ever. This is a massive opportunity for the future of the digital newsroom. And this wave of innovation extends well beyond the newsroom. These advancements are set to redefine video creation across the entertainment industry, advertising, etc., as well. Huge potential ahead for all creators and storytellers. #AI #GenerativeAI #VideoLocalization #NewsTech #TimesNetwork #TimesNetworkNewsroom #FutureOfContent #DigitalMedia

41

1 Comment

A nation that wants to lead in Wave 3–6 AI, robotics, biotech, defense, and energy must identify and optimize high-impact human capital.

1

If you're building with Pipecat, this one is for you. We just launched the Tuner <> Pipecat SDK 𝘱𝘪𝘱 𝘪𝘯𝘴𝘵𝘢𝘭𝘭 𝘵𝘶𝘯𝘦𝘳-𝘱𝘪𝘱𝘦𝘤𝘢𝘵-𝘴𝘥𝘬 Now you can plug in observability and analytics in minutes and get real visibility into how your voice agent is performing in production, at scale. Because shipping the agent is only step one.  Understanding what’s happening in production is where the real work begins. Link in the comments 👇

108

12 Comments

Our team at Agentic Foundation is doing amazing product development work with the Claude Code. - Helping shape the Agentic Code Assistant industry. 🌊 Claude Flow Alpha 80: GitHub-Enhanced Claude Code Hooks. It turns Claude Flow into a fully observable, versioned development layer. Basically the invisible sub agent logic from your Claude Code agents gets recorded directly into GitHub. The new github init command introduces deep GitHub integration with automated checkpointing, instant rollback, and full historical logging of every sub-agent action. It improves introspection by exposing the full execution flow, auditing decisions, tracing context shifts, and analyzing task sequences with timestamped, versioned logs. No more guesswork. Developers get a clear, structured trail of edits, memory updates, and tool activity over time. It also generates a summary-session.md in the .claude/checkpoints/ folder, capturing the full session context for reference or review. This enhancement brings: 🔖 GitHub releases for every checkpoint: edits, tasks, sessions ⏪ Team-wide rollback to any tagged state 🧠 Traceability of what happened, when, and by which agent 🔄 Live syncing across contributors Every hidden operation, task starts, memory changes, context updates is now tracked and published via GitHub. To enable: npx claude-flow@alpha github init --force Ideal for collaborative debugging, multi-agent transparency, and reproducible workflows. 🔗 Full details here: https://lnkd.in/eH69pdDP

89

5 Comments

Strauss Zelnick, CEO of video game developer Take-Two Interactive - "While AI may help with solving something as complex as a cure for a disease or as simple as biology homework, Strauss said that when it comes to creating the sorts of multi-layered universes that Take-Two’s video games have become known for, it’s another story." I will take that any day, if it can cure disease. Video games - let us leave it to humans.

6

Thrilled to announce that Together Fund is investing in Sentra, alongside a16z speedrun! You track results in Jira. Decisions in Notion. Conversations in Slack. But the reasoning, the debates, trade-offs, and context behind why you chose A over B, disappears into what we call "Dark Matter." A decision made in March looks insane by July because no one remembers the constraints that made it smart. I lived this firsthand at Twitter scaling from 800 to 8,000 employees, and at Google while launching AI Overviews to billions at planet scale. The problem isn't process. Process is compensation for something deeper: organizational amnesia. An organization’s "Systems of Record" doesn’t solve this, they encode it. They store what happened, never why. That's why we are investing in Sentra. Sentra is the always-on collective memory that eliminates organizational amnesia by maintaining accurate context for all members and agents, functioning as an operational nervous system. It connects to every channel where work happens, meetings, Slack, email, code commits, docs, calendars, and treats them not as artifacts to search, but as living signals to synthesize. The fleeting and the permanent, unified into a memory that understands. The founding team is built for this: - Jae Gwan Park (CEO): Product-first founder, memory systems research at UofT and MIT - Ashwin Gopinath (CSO): Former MIT professor, created "Reflexion" (NeurIPS 2023), agents that learn from mistakes, 2x founder - Andrey Starenky (CTO): Early Vapi engineer, ex-IBM, built to process enterprise-scale data firehose Together is an operator-led fund. We invest in problems we've lived. This is one of them. Many congrats Jae, Ashwin and Andrey, we are so excited to partner with you! Read the full thesis: https://lnkd.in/gixj9cE4 Book a demo: https://www.sentra.app/ #OrganizationalMemory #AI #Sentra #TogetherFund #a16z #ContextGraphs

71

3 Comments