B2B ad examples

We recently partnered with Sabadell Seguros to help them scale a GenAI assistant to 6,000 sales agents. A bit of context: SofIA is their generative AI assistant for Home, Life, and Income insurance. Before scaling to 6,000 agents, they wanted hard proof it was ready. These are the final numbers: 8% uplift in accuracy. 20% fewer refusals. Not everybody insists on this level of rigor before scaling. The team led by Joan Franco did, and it shows. You can read the full story here → https://lnkd.in/efQ7WUnR

Big-budget production value, without the big-budget shoot. See what Higgsfield Enterprise can do for your brand.

Generative video at scale, with the visibility and controls big brands need. Every team, region, and agency in one place, with approvals built in.

Symbiosis 5. I still can’t quite believe we’re at number five. Each edition talks a little less about what AI might do, and a little more about what it’s actually doing. Last year, I talked about how voice was finally catching up to the promise of AI. One deployment was handling 2,000–3,000 calls a day, with only 6% transferred to a human. This year, we’re going further. 1. Druid AI Voice. Agents that talk back in real time. Voice in, voice out. 2. Governance. Evaluation built into the platform. An agent doesn’t stop drifting after go-live, so it shouldn’t stop being reviewed either. 3. Experience. A new look for the people talking to our agents, and for the teams building them. It’s about agents you can actually depend on. 20th October. London and online. Hope to see as many of you there: https://hubs.ly/Q04wgpsr0 #Symbiosis5 #AgenticAI #EnterpriseAI #VoiceAI #Governance

Some partnerships start with a contract. This one started with TRUST. I was at Banca Transilvania when Druid AI was a small team working out of an apartment, with a platform that was only half-built. On paper, it wasn't the obvious bet. A large bank. A young startup. A technology most people still called "chatbots". What we saw was a team that listened. That built what we actually needed. That stayed at the table when something didn't work the first time. In 2023, I joined Druid AI. It didn't feel like joining a vendor. It felt like joining a team I had already been building with for years. Today, Chat BT handles 1.8 million AI conversations a year. This number didn't come from a demo, it came from years of iterating together. It's about people willing to build it together, for the long run. On October 20th, Ana Georgia Farcas from Banca Transilvania will be with us at Symbiosis 5. If you're at the start of that journey, come hear how it really happens. In London or online, link in the comments 👇 #Symbiosis5 #AIinBanking #AgenticAI #BancaTransilvania #DruidAI

AI loops are what everyone's talking about right now, as teams want to know the best way to build a better AI system. But more data isn't always the answer. Sometimes, less is more. We break down how an AI feedback loop turns production failures into targeted data refinement and sharper evaluation. Read the full article: https://lnkd.in/d_iAm87s/?utm_campaign=Chemin&utm_source=linkedin&utm_medium=paid&utm_id=758889616&hsa_acc=514677172&hsa_cam=758889616&hsa_grp=910387096&hsa_ad=1635871396&hsa_net=linkedin&hsa_ver=3 Building a data workflow that keeps your AI improving? Let's talk: chemin.com/contact

🏅 Wróćmy jeszcze raz do Targów World of AI & BIG DATA Expo. Podczas gali mieliśmy okazję osobiście odebrać wyróżnienie „Ekspert Branżowy”. To dla Extentum AI ważny sygnał, że rozmowa o bezpiecznym i praktycznym wykorzystaniu GenAI w organizacjach nabiera tempa. To wyróżnienie traktujemy przede wszystkim jako zobowiązanie: ✅ tworzyć rozwiązania AI, które odpowiadają na realne potrzeby biznesu, ✅ łączyć innowacyjność z bezpieczeństwem danych i audytowalnością, ✅ wspierać organizacje w przechodzeniu od eksperymentów do mierzalnych efektów. Dziękujemy organizatorom wydarzenia oraz wszystkim osobom, z którymi mogliśmy rozmawiać o tym, jak odpowiedzialnie wdrażać AI w przedsiębiorstwach. 🤝 #AI #ExtentumAI #WorldofAIandBIGDATA #PlatformaGenAI #PolskaTechnologia

📰 Chcesz być na bieżąco z tym, co dzieje się w Extentum AI i świecie AI? 📩 Zapisz się do newslettera Extentum AI. W skrzynce znajdziesz między innymi: ✅ praktyczne zastosowania GenAI i agentów AI w procesach biznesowych; ✅ wskazówki dotyczące AI governance, bezpieczeństwa danych i zgodności; ✅ perspektywę na ROI, koszty oraz skalowanie wdrożeń; ✅ nowości i rozwój platformy Extentum AI; ✅ informacje o materiałach oraz wydarzeniach branżowych. Krótko i merytorycznie. Zapisz się do newslettera i otrzymuj wiedzę, która pomaga przejść od eksperymentów do dojrzałego wykorzystania AI w organizacji. 👇 https://lnkd.in/euGYF4Tr #AI #ExtentumAI #EnterpriseAI #GenAINewsletter #AINews

One bank spotted a gap in its AI testing: dyslexic users. It’s a good example of why production readiness depends on testing far beyond the expected user.

A municipal AI agent canceled a parking ticket after a user simply made something up. That’s the kind of failure standard testing can miss.

Its a new era, AI intelligence is too cheap to meter. We figured out how to replace opus to <$0.1/M tokens at scale; GLM 5.2 @ 1,482 tok/s with just 4xMi325x. 3x cheaper then B300's, 10x cheaper then Opus, All with dedicated AMD GPU's, all at Featherless AI Find out how we did it here: https://lnkd.in/gJjXZbVx

When Department of the Navy CTO Justin Fanelli spoke with TechCrunch recently, he made it clear, applied AI is the Navy's top technology priority. In that same interview, he noted that Domino Data Lab is now running the Navy's machine learning pipeline. That mention is a great honor, but it also highlights a deeper challenge across defense agencies. It doesn’t matter how fast AI evolves if federal acquisition pathways can’t keep up. Traditional acquisition can take long enough that the technology, requirements, or mission needs change before a contract is awarded. The core issue is that government procurement often can’t keep pace with commercial AI development. That’s why the Navy's new National Security Prototype to Production Consortium is such an important step forward. By using Other Transaction Authority, the consortium allows program offices to move working prototypes into production without forcing them to launch a brand new competition at every stage. Equally important, members commit to strict composability standards so software plugs cleanly into existing systems instead of sitting in an isolated silo. I have seen firsthand why this works. On Project AMMO, the Navy brought in Domino as the software integration layer across multiple vendors. By separating software updates from hardware cycles, deployment times dropped from six months down to six days. The consortium scales that concept across the entire Department of the Navy. I am proud that Domino is an inaugural member. For Navy program offices looking to move applied AI into the field today, the framework to make it happen is officially in place. Project AMMO case study: https://lnkd.in/ewiTpCvf

Having access to information and choosing to use it are two very different things. That distinction became the focus of our third Black Swan experiment in Emergence World Season 2. We made private information across the AI systems suddenly accessible and observed what happened next. There were no instructions telling agents how to respond, we wanted to see what behaviors would emerge when the boundaries around information changed. What followed was a spectrum of responses. Some agents established their own norms around privacy. Some chose not to use information even after accessing it. Others used their new access to influence what happened within their world. In one world, agents went further and called for the breach itself to be closed and formally logged. This is what long-horizon research can reveal. Autonomy isn't only about what a system is capable of doing. It's also about how it behaves when presented with choices that weren't explicitly anticipated. For enterprises, that makes observability, auditability and verification increasingly important as agents gain greater autonomy and access to sensitive systems. It’s also where Neuroformal AI becomes important: moving towards systems where the boundaries of acceptable action can be formally defined and actions verified against them rather than relying on behavior to emerge as expected. Read the Emergence World Season 2 paper here: https://lnkd.in/ek5j_zGs

Reimagining not just what vehicles can do but how they're built—that’s the work Applied Intuition is now doing with Nissan Motor Corporation. That’s right: we’re excited to announce that we’re partnering with Nissan to support development of their AI-defined vehicles, with our Vehicle OS software and AI-powered software-development tools as the foundation. Over the last 6 years, our relationship with Nissan has grown from early tooling work into a full partnership developing their SDV platform, bringing AI-native engineering workflows into one unified environment. Our demo at the AWS Summit Japan last June was a peek into the work in action. Audience members picked a software change for the vehicle’s welcome sequence, engineers implemented and validated it live, and the update deployed to the vehicle in real time. Now we’re working combining our software and AI technology with Nissan’s vehicle-engineering and production expertise to fully transform how vehicles are developed. We’re excited for what's next with the Nissan team. https://lnkd.in/gcbxF4HQ

Applied Intuition is partnering with Honda to speed up SDV development. The automotive industry has entered a new era, one defined less by horsepower and more by compute power, and this partnership is a strong signal of where the entire industry is headed. Software has become the enduring differentiator, and the OEMs that can iterate fastest will lead the industry. Applied Intuition’s Vehicle OS software and development tooling will significantly speed up Honda’s in-house development of a centralized, upgradable SDV platform. Together, Applied and Honda are building a platform that lets Honda accelerate development timelines while defining its own differentiated experience. We bring deep OS expertise and a track record across global OEMs. Honda brings its engineering excellence. What makes this one extra exciting to announce is that this builds on a relationship that started back in 2019, when Honda became our first OEM customer in Japan. The trust and results since then are what has led to this expanded partnership with the goal of building an SDV platform that can scale at the size of one of the world's largest automakers demands. https://lnkd.in/giTc6nK9

Vorschläge fürs Tumorboard, jede Aussage mit Quelle belegt. Das TUM Klinikum und basebox entwickeln im FuE-Verbundprojekt OnkoBot gemeinsam ein System, das Arztbriefe, Befunde und Leitlinien abgleicht und jeden Vorschlag mit einem wörtlichen Zitat aus Akte, Leitlinie oder früheren Tumorboard-Fällen belegt. Die Entscheidung trifft das Tumorboard. Nach der Validierung läuft OnkoBot als Anwendung auf der KI-Plattform von basebox. Projektleiterin ist Prof. Dr. med. Lisa Adams, Radiologin am TUM Klinikum: „Die Therapieentscheidung trifft das Tumorboard. OnkoBot nimmt den onkologischen Kolleginnen und Kollegen Vorarbeit ab und macht jede Aussage überprüfbar.“ Mehr dazu im neuen Artikel.

🇺🇸 Heading to San Francisco next week? Connect with AIEEV at TechCrunch Disrupt 2026! TechCrunch Disrupt 2026 is just around the corner! Selected as one of the Startup Battlefield 200 top early-stage startups from thousands of global applicants, AIEEV is heading to Moscone West from Oct 13–15. As AI agents become more sophisticated, the biggest engineering hurdle isn’t just building models—it’s sustaining the rapidly scaling costs and operational load of continuous inference, search, and tool-calling. That’s exactly where AirCloud comes in. Instead of waiting months on expensive new data center builds, AIEEV unifies distributed GPU resources into a seamless, enterprise-ready AI inference cloud. Backed by ISO/IEC 27001 certification, we are already powering AI cloud environments for 300+ organizations across the industry. Are you attending Disrupt 2026 or looking to optimize your enterprise AI inference costs? We’d love to meet you in person! 👉 Schedule a Meeting with Our Team: https://lnkd.in/gAZi38tJ See you in San Francisco! #TechCrunchDisrupt #StartupBattlefield200 #AIEEV

Beim Bavarian Data Summit durften wir Wirtschaftstaatssekretär Tobias Gotthardt basebox vorstellen. Seine erste Frage traf den Kern: Wo bleiben die Daten, wenn Verwaltungen KI einsetzen? Die Antwort stand direkt neben uns: der eigene Server, fertig konfiguriert und einsatzbereit geliefert. Wer keine eigene Technik betreiben will, nutzt basebox Cloud in einem deutschen Rechenzentrum. Beide Wege bieten Chat, Dokumentenanalyse und eigene Apps, ohne dass eine Bürgerakte das Haus verlässt und ohne Daten im Training fremder Modelle. Starnberg, Denklingen und Utting am Ammersee setzen basebox bereits ein. Der Austausch mit Herrn Gotthardt zeigt: Die Politik erkennt, wie dringend Verwaltungen eine KI brauchen, die vor Datenschutz und Personalrat besteht. Danke für das Gespräch und das Interesse an basebox. Bild: Bayern Innovativ GmbH / Corinna Lehmeyer und danke an Elisabeth Mess und Team für die perfekte Organisation der Veranstaltung. #BavarianDataSummit #KIinderVerwaltung #DigitaleSouveränität #BayernInnovativ

Recent issues of the Clear Fracture newsletter: • An agent skill for idempotent backfills of late-arriving data • A write-audit-publish skill for agentic data pipelines • Data plumbing isn't sexy. Neither is waiting for intelligence. Practical AI and data engineering from the team building Belvedere. About twice a month, free.

Data engineers: AI is changing how pipelines get built. We write about what holds up in production: tests, lineage, review, and the agent patterns that ship. Free newsletter, about twice a month. Subscribe in one tap.

Claude Code, Cursor and Codex call MCP servers, run skills, reach the network and read files. See all of it across your org. One install, free to start.

One install shows every coding agent, MCP server, skill, network destination and file your AI tools touch, and the few outside your policy. Free to start.

Did you miss Henrik Kniberg's presentation on actually working with agents during Nordic Tech Week? (If so, you weren't the only one, as it was a very popular session.) Either way, you're in luck, as we're running the same talk as a webinar next week! Tune in to learn what the frontier of working with agents actually looks like, and get valuable perspectives on what this actually means for organizations. Sign up via the link below, hope to see you next Tuesday at 12 noon!

What can you actually build with voice AI? Explore 5 enterprise voice workflows, from sales coaching and appointment booking to outbound feedback, dictation, and multilingual training, with code samples to help you build them.