Category: Technology

  • Enterprise AI is burning tokens without context and teams are paying the price

    PALO ALTO, Calif. May 21, 2026 – DevRev, an AI-native enterprise software company transforming how teams and customers collaborate, today announced the newest release of Computer, by DevRev, its AI teammate for enterprise teams. The release addresses the three biggest failures of modern enterprise AI: models that have no memory of a business, insights that disappear the moment they leave a user’s screen, and AI that answers questions but can’t safely take action.
     
    The rest of the industry has responded to these failures by focusing on selling speed – working faster, producing more outputs, increasing token use – resulting in organizations running world-class models on ineffective infrastructure. The Upwork Research Institute states that 96% of C-suite leaders expect AI to boost worker productivity, but 77% of employees report AI has increased their workload. Focusing on metrics like speed and token volume – “token maxxing” – without clarity distracts from progress, and advocating for applying more compute to inadequate contextual data only produces more noise.The result: knowledge workers lose hours a day hunting for information across disconnected systems, and the people meant to benefit from AI are left holding the bill for its failures.
     
    Computer is built on a different philosophy: delivering both speed and clarity. This solves the context problem through “shared memory,” a curated, living picture of an organization’s data, how the organization works, and how its people interact. Shared memory is the architectural foundation that enables:
    • Precision: answers sourced from real business data, cited and referenced so teams can stand behind them. 
    • Efficiency: trusted answers at lower cost, lower token usage, no analyst required, full context on the first response. 
    • Safety: nothing goes out before a human approves it, with full audit trails and the ability to undo any agent action. 
     
    When AI has this context, it stops guessing and starts acting like a trusted member of the team. Reliable performance builds trust; trust unlocks action. That progression – from trusted answers to safe actions – is what makes multiplayer collaboration between humans and AI possible across every function.
     
    “Every AI company is selling speed, which is fueling the next enterprise crisis. This will be the latest form of workplace burnout if we don’t do something about it,” said Dheeraj Pandey, Co-founder and CEO at DevRev. “Speed without the right context is just faster noise, noise that overloads humans in the loop, and eventually breaks them. Computer is built on a different philosophy: work softer. Give AI the enterprise memory and shared context it needs to perform reliably, and then let it take action. Only then will your people have the confidence that the AI they use is performing the tasks they need it to, more accurately and with less handholding.”
     
    What is new in the newest release of Computer
    The evolution of shared memory at every level ensures that every Computer session now builds on the last. At the individual level, Computer learns how each person works, picking up where they left off with each new session. At the team level, the skills and AI agents one person develops become available to everyone. At the organizational level, institutional knowledge stays in the system permanently. When a top-performing rep leaves, their account knowledge does not leave with them.
     
    Before this release, AI insight disappeared the moment it left one person’s screen. Now the introduction of Multiplayer AI lets teams share a live Computer session where everyone sees the full context and continues the analysis together. Colleagues can question, build on, and correct reasoning in real time. In a 2025 study, KPMG and University of Melbourne reported that 57% of employees admit to using AI in non-transparent ways, including avoiding revealing when they have used AI tools to complete their work. Teams rarely benefit from one another’s AI work. Multiplayer AI changes the unit of attribution from “what I did with AI” to “what we did with AI, together.”
     
    Earlier versions of Computer answered questions and took single-step actions. The addition of the new desktop app shifts Computer from a question-answering tool to a content-producing system. From the in-app canvas, any user can generate complete, fully branded and formatted work artifacts grounded in real business data: competitive slide decks, QBR reports, structured dashboards, knowledge base articles, and multi-step workflows; these outputs are available in a variety of file formats, including PPT, HTML PDF, DOCX, and more. Skills and outputs built in the Canvas are saved at the user, team, or organization level and become reusable across the business.
     
    Agent Studio gives any team the ability to build, test in a sandbox environment, and deploy AI agents that take action across connected systems. Every action runs under individual user permissions, not a shared account. Every step is traceable, auditable, and reversible: if an agent makes a mistake, it can be rolled back.
     
    Key capabilities at a glance
    • Shared Memory: personal, team, and organizational memory that compounds over time
    • Trusted Answers: intent-aware search and consistent data answers. Same question, same answer, every time. Computer doesn’t guess – it knows, and it shows its work.
    • Safe Actions: governed, auditable actions across your systems. Computer acts on your behalf with guardrails, so teams move faster without risk.
    • Multiplayer AI: shared live sessions for human-to-human, human-to-AI, and team-wide collaboration
    • Skills – reusable workflows that any team can build, share, and deploy. One person’s expertise becomes everyone’s capability. Agent Studio allows users to build and deploy AI agents with sandboxed testing, full audit trails, and rollback
     
    Additional capabilities:
    • Text2SQL: analytical queries across structured data in plain language, no data analyst required
    • Connectors: Gmail, Outlook, Slack, Notion, Google Drive, Jira, Microsoft OneDrive, SharePoint, and any MCP-compatible tool
    • Usage-based pricing: that scales with adoption and not headcount
     
    With the newest release of Computer, the above capabilities are all generally available today.
     
    Results from customers in production
    More than 250 organizations have Computer live in production, with over 1,000 users onboarded since launched in September 2025. Customers include BILL, HDFC Bank, and FAME, spanning financial services, aviation, retail, and technology.
     
    The results Computer, by DevRev’s customers are reporting speak for themselves. Customers Pebl and Uniphore are resolving 85% of support tickets without any human involvement. BILL has achieved around $5M in operational savings. India’s largest airline went from kickoff to production in 14 days and selected Computer over Salesforce Agentforce in a head-to-head evaluation. A retail loyalty customer is saving $1.2M annually, with sales reps reclaiming six hours a week and the team reporting a 30% productivity boost. And FAME is saving users more than 10 hours a week, resolving tickets approximately 40% faster, and accelerating specific workflows by up to 75%.
     
    “Customer experience is a top priority for us, and AI presented a real opportunity to set a new standard while reducing cost,” said Steve Januario, CIO at BILL. “With DevRev, we’re seeing how agentic AI can actually reduce support costs and help customers get the answers they need faster, without compromising on quality.”
     
    Availability
    • Across web, mobile, and desktop. Use Computer from any device.
    • Inside the DevRev system of record apps. Existing DevRev customers can access Computer directly within their current workflow without switching surfaces.
     
    Resources
    Visit devrev.ai to learn more about the enterprise offering, book a demo, or start a free trial of Computer. 
     
    Usage-based pricing with plan and billing management are available at https://devrev.ai/pricing
  • AAEON’s BOXER-8629AI Recognized for its Innovation and Market Potential with Best Choice Award

    AAEON’s BOXER-8629AI Recognized for its Innovation and Market Potential with Best Choice Award

    With an IP67-rating and support for four GMSL2 cameras, AAEON’s BOXER-8629AI has been recognized as a standout due to its potential to resolve deployment challenges across diverse environments.

     

    Taipei, Taiwan – May 21 – Leading edge AI platform provider AAEON (Stock Code: 6579) is proud to announce that it has been honored with the Category Award for the BOXER-8629AI at the COMPUTEX Best Choice Award. The awards ceremony is due to take place during COMPUTEX Taipei, held from June 2nd to June 5th at Taipei Nangang Exhibition Center.

    The BOXER-8629AI is a rugged, compact, IP67-rated advanced fanless embedded AI system powered by NVIDIA Jetson Orin Nano with DRAM directly on-module and high optimized memory efficiency for AI workloads. Despite measuring just 135mm x 120mm x 87.5mm, the system is equipped with a broad selection of features conducive to integration in autonomous vehicle, smart city, and other emerging sector applications. Central to this is the BOXER-8629AI’s support for four GMSL2 cameras for high-bandwidth, low-latency video acquisition, GNSS, 9-axis sensor, and IMU module support, as well as native CAN FD for actuator and motor control.

    The product’s market potential is augmented by its rugged design, key for deployment in harsh industrial and outdoor environments. The BOXER-8629AI is IP67-rated, with its external I/O ports featuring O-ring sealing to the same standard to prevent water and dust ingress during operation.

    AAEON has also highlighted the system’s modular design as a benefit to integrators, with custom I/O panels available for the purpose of meeting different project requirements. Meanwhile, the system is equipped with a broad 12V to 24V power input range with built-in ignition control and protection circuits, as well as military-grade vibration, shock, and drop resistance.

    AAEON is honored to receive the award, particularly given the prestige of being highlighted by a platform that has been central to recognizing the functionality, innovation and market potential of new technological solutions for over 20 years. COMPUTEX, Asia’s largest annual ICT exhibition, plays a key role in showcasing the very best products that industry leading companies have to offer, providing a one-stop procurement platform for global buyers across industry segments.

    The BOXER-8629AI will be on display at both the AAEON COMPUTEX booth (#J1310, 1st Floor, Hall 1), as well as the award-winning products display area (4th Floor, Hall 2) throughout the duration of the event.

  • AVer HUB Series Now Includes 5-Year Warranty, Offering Lasting Support for Hybrid Workspaces

    AVer

    Taipei, Taiwan – May 21: AVer Information Inc., an award-winning provider of AI audio-video solutions, announces the extension of its industry-leading 5-year warranty to the HUB Series, including the HUB30 and HUB35-EXT35 4K BYOM Matrix Switchers. This underscores AVer’s commitment to providing organizations with long-term reliability and peace of mind for their essential meeting room infrastructure.

    AVer understands that managing multiple meeting room devices can be a complex challenge and is committed to making these connections as seamless and efficient as possible. By providing a 5-year warranty, AVer ensures that the HUB Series, serving as the command center for the entire meeting space, is protected for the long haul. For the average user, this means fewer technical disruptions and the confidence that their meeting setup is built to last.

    The HUB Series backed by this 5-year warranty can streamline device management in virtually every collaborative environment. Organizations in fields such as corporate business, hybrid education, and live event production can elevate their meeting spaces with these new 4K BYOM Matrix Switchers, which offer the following benefits:

    ŸSmart Multi-Camera Switching (ConnectAI): The HUB Series enables seamless switching between two AVer cameras, using AI-driven algorithms for dynamic, multi-angle video in live streams, hybrid meetings, and video productions.

    ŸCrystal-Clear Visuals: Video is delivered in 4K UHD resolution, ensuring exceptional visual clarity.

    ŸCentralized Presentation Control: Supporting the CP10 G2 Touch Panel, a single user-friendly interface, presenters can effortlessly configure dual 4K displays for flexible viewing.

    ŸBYOM-Ready Connectivity: These switchers offer versatile connection options — including HDMI, USB-C, and DisplayPort — while powering personal devices reliably during long sessions.

    ŸFlexible Scalability: Whether using the plug-and-play HUB30 for standard meetings or the HUB35-EXT35’s included 4K USB extender for larger spaces like auditoriums, users maintain flawless signal integrity across any distance.

    Originally introduced in 2023 for AVer’s Pro AV cameras and MT Series Matrix Tracking Boxes, this industry-leading 5-year warranty now brings that same long-term confidence to the expanding AVer Pro AV lineup. For further information about the warranty, please reach out to AVer or your local authorized dealer or distributor.

     

     

  • Silver Trak Digital and Media Room Connect win IAMT Best Products and Services Award

    At BroadcastAsia, 20-22 May at Singapore Expo on IAMT Pavilion, on stand 5A3-3

     SINGAPORE, 21 May 2026 – Media asset management specialists Silver Trak Digital have won the IAMT Best Products and Services Award for their Media Room Connect solution at BroadcastAsia 2026. Media Room Connect allows users to create and manage secure, branded screening rooms directly on their company domain.

    It is a key part of the Media Room Media Asset Management solution which enables secure sharing, global content delivery and screener campaigns for broadcast and production assets.

    Silver Trak Digital and Media Room Connect

     

    (l-r) IAMT CEO Saleha Williams and Silver Trak Digital COO Christian Christiansen 

    IAMT CEO Saleha Williams said, “A huge congratulations to Silver Trak Digital for winning the IAMT Best Products and Services Award with Media Room Connect. Celebrating those driving changes across the APAC region, the IAMT Awards are judged by an independent panel of industry experts and recognises standout achievement, innovation and industry impact.”

    Silver Trak Digital is demonstrating its next gen Media Room Suite solutions at Broadcast Asia 2026 from 20-22 May at the Singapore Expo on the IAMT Pavilion, stand 5A3-3.

    Silver Trak Asia with their IAMT Award

     

    (l-r) Christian Christiansen and Karlz Varma, VP Business Development for Silver Trak Asia with their IAMT Award

    Silver Trak COO Christian Christiansen said, “We are delighted and honoured to have won the IAMT Best Products and Services Award for Media Room Connect, which is a product that has been developed very much with industry requirements in mind. Our thanks go to the IAMT, the awards’ judges and our clients who have helped make Media Room and Media Room connect such a success across the Asia Pacific region.”

    Silver Trak Digital are demonstrating two products from their latest, cutting-edge Media Room suite at BroadcastAsia this year – Media Room covering Media Asset Management, distribution and screeners (and more) and Media Room Connect which creates and manages a secure screening room on your company domain.

    They are also discussing two new products under development including Media Room Screeners, a secure screener campaigns product with enhanced artwork capabilities and Media Room Transport, our SaaS large file, high speed transport solution.

    Media Room is the affordable, secure, cloud-based media asset management and delivery platform that enables content owners to upload, manage, market and distribute content globally, faster, more efficiently and with complete control.

    Due to its flexibility and affordability, media organisations can easily start small and scale with Media Room as their business grows.

    To get a demo of Media Room Connect see Silver Trak Digital at BroadcastAsia 2026 on the IAMT Pavilion, stand 5A3-3.

  • Struddys Apparel becomes first global site to install new Epson SureColor F20060

    BRISBANE, May 21 – Struddys Apparel, a manufacturer specialising in custom teamwear, sportswear and branded garments, has become the first site and company globally to purchase and install the new Epson SureColor F20060 dye sublimation platform, strengthening its position as one of Australia’s leading vertically integrated apparel producers. 

    Struddys Apparel becomes first global site to install new Epson SureColor F20060

     

    The investment represents the latest step in Struddys’ growth from a small custom operation into a large-scale, Australian-made production facility with a strong focus on quality, speed and full in-house control. It also highlights how strategic technology adoption is helping local manufacturers scale output while maintaining the flexibility needed for custom and short-run work.

    struddys

    Struddys has been a strong example of organic growth through dye sublimation. The business began with a single 64-inch printer (delivering approximately 20 sqm/hr) and expanded to three units as customer demand increased. As production volumes grew, however, the company reached the practical limits of adding smaller-format devices.

    At that point, Struddys made a deliberate move to higher-volume production capability rather than continuing to scale incrementally. The company adopted a 76-inch Epson SureColor F10060, a production platform capable of up to 255 sqm/hr and specifically 100 sqm/hr in apparel production, enabling a major increase in throughput and operational efficiency.

    That shift gave Struddys a more effective production model: high-volume work could be concentrated on the larger-format platform, while its 64-inch printers remained ideal for short runs, custom programs and bespoke jobs.

    As demand has continued to grow, Struddys has now added a second 76-inch platform with the purchase of the new SureColor F20060.

    Struddys 2

     

    The new installation makes Struddys the first global site to adopt the Epson SureColor F20060 and positions the company at the forefront of dye-sublimation production capability.

    The additional capacity will support faster turnaround times, greater production resilience and further growth as the business expands across teamwear, sportswear and branded apparel markets.

     

    signature_3316152233

     

    The Epson SureColor F20060 is a dye-sublimation printer designed for textile manufacturers and print service providers requiring fast turnaround times. Built for demanding production environments, it combines consistent output quality, reliable operation and printing speeds of up to 300 sqm/h. It is suited to industrial-scale production including fashion apparel, sportswear, home furnishings and soft signage.

    For Struddys, the addition of the F20060 is about more than just speed alone. It supports a production strategy that combines scale with agility allowing the business to meet larger volume orders efficiently while preserving the responsiveness needed for custom, bespoke and short-run apparel work.

    This balance is increasingly important for manufacturers serving customers who expect both rapid delivery and tailored product offerings.

    By bringing the F20060 into its operation, Struddys is reinforcing a model built on local manufacturing, controlled workflows and production flexibility. With Epson technology underpinning both its high-volume and specialty output, the company is well placed to continue growing without compromising the quality and service standards that have driven its success.

    Struddys’ journey from a single-printer operation to a scaled production business illustrates how investment in the right technology at the right time can transform manufacturing capability. Its latest move, becoming the first global site to install the Epson SureColor F20060, signals not only confidence in future demand, but a clear commitment to staying ahead of the curve in Australian apparel production.

  • Averlon Launches Precog to Stop Exploitable Risk Before It Reaches Production

    New predictive remediation capability helps security teams prevent exploitable risk before it reaches production

     

    Redmond, WA, May 21, 2026 — Averlon today announced Precog, a predictive remediation capability that identifies exploitable risk in proposed code and infrastructure changes and delivers the fix to developers before the change reaches production. Precog addresses a widening gap: AI is accelerating both code delivery and vulnerability discovery, and security teams can no longer manage risk only after it lands in production. 

    The need for this shift is becoming urgent. Google Cloud’s Mandiant M-Trends 2026 report found that mean time to exploit collapsed from 63 days in 2018 to an estimated minus seven days in 2025, meaning exploitation now often begins before a patch is available. New frontier models such as Claude Mythos and GPT-5.5-Cyber are making it increasingly clear that AI will compress the time required to discover, validate, and exploit vulnerabilities. The result is a widening gap between the speed at which risk is discovered and exploited, and the speed at which security teams can triage and fix it.

    The industry is converging on a new operating model: Remediation Operations, or RemOps. The premise is simple: finding risk and closing risk are different problems. Security teams do not need more alerts; they need a way to understand what is truly exploitable, prioritize by business impact, and drive safe fixes through developer workflows. 

    Averlon’s Remediation Operations platform addresses the full lifecycle of risk reduction: ingesting security findings, determining what is truly exploitable, prioritizing by business impact, and driving agentic remediation through developer workflows. The platform has helped customers reduce remediation time by up to 90 percent and alert noise by up to 95 percent, helping security teams move from backlogs of thousands of findings to the handful that need fixing. 

    With Precog, Averlon extends that model earlier in the lifecycle by preventing exploitable risk before it becomes production exposure. Unlike security scanners that flag findings based on generic severity scores, Precog evaluates whether a proposed change would actually be exploitable in the customer’s real environment, accounting for internet reachability, exposed services, and existing compensating controls. This contextual analysis means Precog surfaces the changes that genuinely create exposure, not the long tail of theoretically risky findings that wouldn’t be exploitable in production. Precog integrates into CI systems such as GitHub, evaluating proposed changes before they reach production. 

    When risky changes are detected, Precog identifies the issue, explains the exploitable path, and generates a remediation directly in the developer workflow. Developers receive the proposed fix at the same time they are notified of the risk, reducing friction between security review and software delivery. 

    “AI is changing both sides of the software lifecycle. It is accelerating development while also introducing code that is often not ready for production. With AI also accelerating the discovery and exploitation of weaknesses in that software, security teams can no longer rely only on post-production detection and backlog management. Capabilities like Averlon’s Precog point to where the market is headed: identifying risky changes earlier and helping developers fix them before they become production exposure.” 

    Chris Steffen, VP Research, Enterprise Management Associates

    Complementing Precog is Vulnerability Intelligence, Averlon’s CVE research feed at research.averlon.ai, which gives teams the context they need to triage emerging vulnerabilities, including exploitability, attacker requirements, required privileges, user interaction, and evidence of exploitation in the wild. 

    “Security teams have relied on finding and fixing vulnerabilities after they reach production. AI has made that untenable from both directions: it is generating new vulnerabilities faster than teams can triage them, and it is collapsing the window between exposure and exploitation. You cannot remediate your way out of that. The only way to stay ahead is to prevent exploitable risk from reaching production in the first place. That is what Precog does.”

    Sunil Gottumukkala, CEO, Averlon 

    Learn more and see Precog in action: 

    ·         Precog: Preventing Exposure Before the Backlog (blog): averlon.ai/blog/averlon-precog-preventing-exposure-before-backlog

    ·         Vulnerability Remediation: Down and Up We Go (blog): averlon.ai/blog/vulnerability-remediation-down-and-up-we-go

    ·         Vulnerability Intelligence Brief #1 (research): averlon.ai/blog/vulnerability-intelligence-brief-cve-2025-11953-react-native-metro-command-injection

    ·         Vulnerability Intelligence (live research feed): research.averlon.ai

    ·         RemOps End-to-End Demo (recording): averlon.ai/webinars/remediation-ops-live-demo

    ·         Averlon Platform: averlon.ai/platform

    Averlon is pioneering Remediation Operations, helping security teams move from vulnerability findings to safe, prioritized fixes. Its agentic AI identifies what is truly exploitable and automates remediation directly within developer workflows. Averlon is trusted by security teams at enterprises and fast-growing companies to efficiently scale their security programs. Founded in 2022, Averlon is backed by Salesforce Ventures, Voyager Capital, and Outpost Ventures.

  • hubergroup introduces advanced resin technology for offset inks

    hubergroup

    hubergroup Print Solutions is unveiling another major innovation designed to better meet today’s market requirements. After intensive development work, the international printing ink specialist is launching the next generation of its conventional offset inks, featuring a completely new resin technology. The advanced formulation delivers measurable improvements in printing performance while maintaining consistently high print quality. Printing companies thus benefit from significantly fewer make-ready sheets, more consistent colour reproduction, and a substantial reduction in misting – even on high‑speed presses. The relaunch covers well‑established commercial and packaging ink series, including RESISTA, RAPIDA ECO, ECO-PERFECT-DRY, PACKAGING PLUS, MGA NATURA, and MGA CORONA.

    Why resin technology matters

    Resins form the backbone of offset ink systems, defining fundamental characteristics – from physical properties such as viscosity or tack to print performance factors like misting resistance, setting speed or adhesion. Recognising their central role, hubergroup invested in research and development to engineer a resin technology that enhances these properties across the board.

    „An improved resin means an improved ink,” explains Richard Gill, Product Manager Sheetfed at hubergroup. “As we at hubergroup manufacture our own resins, we can tailor the formulation to precisely meet the performance characteristics we want. This capability enabled us to develop a perfectly aligned resin technology for our conventional offset inks.”

    Real advantages in everyday production

    During extensive long‑term production testing, the advanced resin technology has demonstrated clear, quantifiable benefits:

    Measurable reduction in misting
    Wider water window for more stable printing
    Significantly lower dampening levels
    Fewer sheets required for on‑press adjustments
    Stable print density and consistent colour results

    At the same time, post-print performance remained consistently good.

    Global rollout across the offset portfolio

    hubergroup will introduce the new resin technology across its entire global conventional offset portfolio during the first half of 2026. Product names and sales codes will remain unchanged, ensuring a smooth transition.

    “With this new technology, we are one step ahead of the market and fully aligned with current and future needs of print and packaging companies,” summarises Richard Gill. “We invite customers to reach out to us at any time to learn more about the advantages of this innovation.”

     
     
  • Caspia Technologies Significantly Expands Agentic Silicon Security Verification with New Release of CODAx

    GAINESVILLE, Fla., May 20: Caspia Technologies announced a significant expansion of its flagship security verification product CODAx. New features were described along with the impact on agentic workflows and associated customer momentum. Some details of Caspia’s security audit of the open-source Caliptra root-of- trust were also disclosed. And the company provided a preview of its communication plans over the next two months.

    CODAx, Caspia’s RTL security analyser, was updated with two new AI- driven features, Asset Assist and Report Assist. These features significantly enhance the usability for wider deployment across all design teams, without requiring security expertise.

    Security analysis begins with understanding what needs to be protected. Current practices in identifying security assets are manual and require expert assistance. This results in limited identification of security assets and limited scope and coverage for security analysis.

    Asset Assist performs automated asset identification on the target design using design-specific details to find a much broader list of security-relevant attributes. The result is a more targeted and comprehensive security analysis. Examples of security assets include untrusted signals, protected states, and critical control signals.

    Report Assist addresses the process of understanding the impact of violations to develop a prioritised solution strategy. Current practices can be extremely time-consuming and also require expert assistance. Report Assist delivers AI-driven analysis of CODAx security violations with context- driven insights, design-specific severity ratings, and tailored impact scores. This automated violation triage delivers much faster and more effective strategies to address key issues.

    These new features can reduce weeks to months of security verification to minutes to hours, allowing existing design teams to perform advanced security verification at scale.

    “Report Assist and Asset Assist are paving the way for automated, agentic security verification,” said Stuart Audley, VP and GM of product management at Caspia. “We are seeing significant traction across first-tier CPU, GPU, and memory providers. I am delighted to see our agentic vision begin to take shape.”

    The Caliptra root-of-trust represents a production-grade, open-source security IP developed to serve as a foundational trust anchor in SoC designs. It contains over 700 design files, amounting to more than 300,000 lines of RTL code.

    The analysis identified 10 critical items and over 50 high-priority findings. These vulnerabilities are exactly the kind of footholds an attacker needs to wedge into a device and take control.

    Caspia shared the details of errors associated with one directive related to fault injection susceptibility for sensitive controller states. This directive analyzes the fault injection feasibility metric for finite state machine (FSM) state encodings. Caspia explained that violating this directive makes glitches or fault-induced jumps into protected behavior more feasible, impacting integrity and resulting in potential information leakage.

    “The Caliptra design sees widespread use in the industry,” said Beau Bakken, VP of products at Caspia. “We are working with the appropriate contacts in the industry to share these results and assist in developing appropriate corrective action.”

    On May 22, 2026, a podcast with Stuart Audley will be released that discusses agentic workflows for security verification on the Semiconductor Insiders podcast series. On June 24, 2026, SemiWiki.com and Caspia will present a webinar entitled Scaling Security Verification Across Design Teams with CODAx Static Analyzer. Beau Bakken and Dr. Zahin Ibnat will present. And on July 27-29, 2026, Caspia will exhibit at the 63rd DAC Chips to Systems Conference in Long Beach, CA. Visit Caspia in booth 759.

  • AVIAN raises $2.6M to scale AI thermal monitoring for the world’s most fire-prone industrial sites

    After two profitable bootstrapped years, the Zurich-based industrial AI company will accelerate deployment of its end-to-end thermal monitoring platform across sawmills, recycling, mining, chemical processing, oil and gas, and maritime operations – categories where fire and downtime risk are increasingly outpacing what insurers are willing to cover.

    Zurich, Switzerland — May 20 ; Industrial operators across Europe and North America are facing a problem they can’t inspect their way out of. Fine dust, friction, electrical faults, and aging equipment are pushing fire and downtime risk into territory that insurers will no longer underwrite at viable premiums. Sites that were insurable five years ago are being deemed too risky today. Against that backdrop, AVIAN – the Zurich-based industrial AI company building 24/7 thermal monitoring for the world’s most fire-prone facilities –  today announced it has raised a $2.6M pre-seed round, led by Founderful. 

    AVIAN was profitable and entirely bootstrapped for two years before raising. The company says it raised this round to go faster, expand engineering and deployment capacity, and scale beyond its stronghold in wood products into recycling, chemical processing, oil and gas, and maritime. AVIAN is on track to surpass $1M in ARR in 2026.

    How AVIAN works 

    Insurance markets are tightening, and more sites are being treated as high risk as equipment ages and failure rates climb. The old approach to thermal safety still looks like periodic thermography, a technician walking the floor with a handheld camera once a quarter. That method misses the window that matters: the hours when a component starts running hot before it fails. Most thermal vendors also stop at the hardware, selling a camera and leaving operators to figure out setup, monitoring, and escalation on their own. AVIAN takes a different approach: the sensor is one component of solving the problem, not the product. Customers are typically up and running in minutes, not months.

    AVIAN is built to run like an always-on reliability layer. Its thermal cameras continuously watch the critical components that most often become ignition points, motors, bearings, conveyors, presses, and electrical cabinets, and learn what “normal” looks like in that specific plant. From there, the system focuses on drift, the early heat patterns that show up before failure. Smart alarms filter out routine heat sources so teams aren’t chasing noise, and alerts go to the right people with enough lead time to intervene before a hot component turns into downtime or fire.

    AVIAN also generates automated predictive maintenance reports and backs the platform with 24/7 human support. Every alarm event is reviewed and fed back into the models, so detection keeps improving across the fleet and each new site benefits from what AVIAN has already learned in the field.

    Customer results

    Over the last two years, AVIAN has prevented $50M+ in damages from fires and equipment failures and is deployed in approximately 50 sites across 9 countries.

    The impact is clear. Kamps Pallet reduced annual insurance costs by 10% at its Dillwyn sawmill after deploying AVIAN’s system. Sierra Pacific Industries has avoided 24+ hours of unplanned downtime at its Quincy site in the last 12 months alone. Schilliger Holz has used AVIAN to avoid fires and run tighter against unscheduled stops.

    In several cases, the system has caught incidents at the point where they still looked small. A pellet press fire was detected early for a customer in Switzerland, avoiding millions in damage. In Germany, AVIAN flagged a small electrical fire next to a machine worth millions. Containing it early protected both the asset and the next 6 to 18 months of production that could have been lost waiting for a replacement.

    “AVIAN has developed a solution to a problem which probably affects everyone in the industry directly. For us, it is a great partnership as it helps us make our operations much safer and improves the monitoring process. You will never be able to reduce the risk of fires to zero, but you can do everything you can to minimize the danger as much as possible – and AVIAN makes that possible in a simple and straightforward way,” said Ernest Schilliger, CEO Schilliger Holz. 

    The team

    AVIAN is a 10-person team based in Zurich. The company was founded after one of Switzerland’s largest sawmills saw Hanover’s robotics and AI research in the Swiss media and reached out about escalating fires, downtime, and rising insurance pressure.

    “Most operators don’t need another camera. At 3 a.m., they need to know that a bearing is running hot before it ignites the dust around it,” said Drew Hanover, Co-Founder and CTO of AVIAN. “We bootstrapped the business for two years because we wanted to build something operators actually trusted. We raised with Founderful for one reason: to keep doing that, in more markets, faster, without changing what we are. We spent zero minutes on a deck.”

    Alex Stöckl, Partner at Founderful added: “Within a year of incorporation, the team at AVIAN already served dozens of manufacturing businesses in the US and Europe, preventing real fire incidents on a daily basis. With their thermal-vision technology, there’s an immediate ROI and a new industrial intelligence layer that unlocks further use cases and value for customers over time – backing them to accelerate their go-to-market and product roadmap was a no-brainer.”

    What’s next

    AVIAN’s roadmap splits into two tracks. First, the company has spent years building relationships with insurers to understand how risk is assessed, and its growing camera fleet is positioned to produce something underwriters increasingly demand: real-time, site-level risk assessments backed by live thermal telemetry. Second, AVIAN Vision extends the platform beyond thermal by upgrading existing CCTV systems to detect smoke and fire, giving customers a way to broaden 24/7 protection across an entire facility without replacing the infrastructure they already operate.

    AVIAN’s long-term thesis is straightforward. Industrial risk has been priced for decades using actuarial tables and historical claims data. The next decade will be priced using quantifiable, real-time operational data, and AVIAN believes sites that have become difficult to insure can become insurable again through data, discipline, and proactive systems.

     

  • AAEON’s EPIC-BTS9 Delivers Intel Core Series 2 Processing on a 4″ Single-Board

    AAEON’s EPIC-BTS9 Delivers Intel Core Series 2 Processing on a 4

     

    With up to 65W socket-type CPU support, quad-LAN, and eight-lane PCIe Gen 4 expansion, AAEON’s EPIC-BTS9 is positioned as a foundation for smart manufacturing and industrial robotics.

     

    (Taipei, Taiwan – May 20, 2026) AAEON (Stock Code: 6579), a leader in industrial PC solutions, has released the EPIC-BTS9, the latest addition to its 4″ EPIC Board portfolio, making it the company’s first single-board to support up to 65W socket-type CPUs from the new Intel Core Series 2 socket-type processors (formerly Bartlett Lake).

    While primarily developed for use with Intel Core Series 2 processor CPUs, which grant up to 24 cores and 32 threads of processing performance, the EPIC-BTS9 is also pin-to-pin compatible with 12th, 13th, and 14th Gen Intel Core processors, up to 65W.

    In addition to its cross-generation CPU compatibility, the EPIC-BTS9 is available in SKUs equipped with a choice Intel R680E, Q670E, or H610E chipsets, with the board’s Intel R680E Chipset SKU offering ECC (Error-Correcting Code) support to the 64GB of dual-channel DDR5 system memory on offer via two SODIMM slots.

    The EPIC-BTS9’s I/O comes with a total of four LAN ports, three running at 2.5GbE with hardware timestamping, and one at 1GbE speed. These are joined by two USB Type-A ports offering USB 3.2 Gen 1 and Gen 2 signals, respectively. Rounding off its physical I/O is a HDMI 2.0 port.

    AAEON’s EPIC-BTS9 Delivers Intel Core Series 2 Processing on a 4

     

    For internal connectors, AAEON’s positioning of the board for industrial AI and smart manufacturing makes sense, with dual COM connectors for RS-232/422/485, an 8-bit GPIO, and SMBus. I2C is also available when running Windows, rather than Ubuntu, and can be changed by BOM. The board also offers headers for both VGA and eDP 1.4, the reasoning for which being to allow simplified integration with existing industrial display panel setups.

    The EPIC-BTS9 offers versatile expansion options, with the main standout being an eight-lane PCIe Gen 4 slot, which can be used for AI accelerators, graphic capture cards, or additional storage needs. Meanwhile, the board hosts M.2 2280 M-Key, M.2 3052 B-Key, and M.2 2230 E-Key slot for additional storage and wireless communication module installation.

    The EPIC-BTS9 is now in mass production, with samples available for order on the AAEON eShop.