How AI is Transforming Connectivity

Agent to Agent communication between software will be the biggest unlock of AI. Right now most AI products are limited to what they know, what they index from other systems in a clunky way, or what existing APIs they interact with. The future will be systems that can talk to each other via their Agents. A Salesforce Agent will pull data from a Box Agent, a ServiceNow Agent will orchestrate a workflow between Agents from different SaaS products. And so on. We know that any given AI system can only know so much about any given topic. The proprietary data most for most tasks or workflows is often housed in many multiple apps that one AI Agent needs access to. Today, the de facto model of software integrations in AI is one primary AI Agent interacting with the APIs of another system. This is a great model, and we will see 1,000X growth of API usage like this in the future. But it also means the agentic logic is assumed to all roll into the first system. This runs into challenges when the second system can deliver a far wider range of processing the request than the first Agent can anticipate. This is where Agent to Agent communication comes in. One Agent will do a handshake with another Agent and ask that Agent to complete whatever tasks it’s looking for. That second Agent goes off and does some busy work in its system and then returns with a response to the first system. That first agent then synthesizes the answers and data as appropriate for the task it was trying to accomplish. Unsurprisingly, this is how work already happens today in an analog format. Now, as an industry, we have plenty to work out of course. Firstly, we need better understanding of what any given Agent is capable of and what kind of tasks you can send to it. Latency will also be a huge challenge, as one request from the primary AI Agent will fan out to other Agents, and you will wait on those other systems to process their agentic workflows (over time this just gets solved with cheaper and faster AI). And we also have to figure out seamless auth between Agents and other ways of communicating on behalf of the user. Solving this is going to lead to an incredible amount of growth of AI Agents in the future. We’re working on this right now at Box with many partners, and excited to keep sharing how it all comes evolves.

In the past two weeks, I’ve met with leadership at NiCE, Amazon Web Services (AWS), Zoom, and 8x8. There’s a pattern - here are some reflections on these conversations. AI is eating the enterprise communications playbook for breakfast. If you're not paying attention, you're already behind. Key assumptions that have guided the industry for decades are rapidly becoming obsolete in the age of AI. Here are SIX critical shifts occurring: ONE: Voice is the New UI. Remember “My voice is my passport” as a security phrase, now it’s my voice is my keyboard. APIs are old school. Enterprise-wide, applications and integrations will be voice-enabled from meetings (with AI scribes) to customer service. The future is frictionless, voice-first interactions and integrations, multilingual, and without code. TWO: Mind Your Data: AI without contextual data is like a kiss without a squeeze. Every data repository is a treasure trove, and new moats protect the repositories instead of business practices. Examples include Microsoft putting up CAPTCHA to access Teams meetings and Slack locking down its customers’ data. The new browser wars are unconcerned about eyeballs. THREE: Workflows are the New Apps. Forget simple automation. We're entering the era of AI-native automation, where AI handles complex workflows that require judgment. An AI that doesn't just listen to a customer call but understands the intent, updates the backend systems, and routes the follow-up autonomously. #GameChanger. FOUR: Bottlenecks Be Gone: The modern workplace has largely been throttled by human bottlenecks, and these bottlenecks will disappear. We see this first with code generation; developing new code is becoming the fastest part of a project. Other bottlenecks are various barriers to decisions, such as data collection and analysis. We are moving from concept to code to scale in days, not months or years. FIVE: Soon This Will Matter: As disruptive and consuming as AI has become, none of this matters, yet! That’s because AI isn’t that useful, yet. We are in the Scantron era again. Scantron bubbles revolutionized paper scoring. A good step, but digitization is what mattered. AI is automating existing workflows. The real stuff comes in the reimagination of work. The first glimpse of this is in agentic AI. Focus on outcomes, not processes. SIX: The Barriers to Entry are Changing: The comms sector has enjoyed numerous barriers to entry over the decades, and most of them are disappearing. AI is simultaneously commoditizing and enabling competitive advantage. Giants may or may not fall, but their businesses will radically change. The Giants Cometh. #AI #FutureOfWork #EnterpriseCommunications #VoiceAI #Automation #Tech  #UCaaS #CCaaS Tanya (Blackburn) Shuckhart John Sun Megan Donaldson Schevone Johnson

Software systems work best when they’re connected to each other. For years, incumbents use deep integrations as a competitive moat. But AI upends this dynamic. A few of our portfolio companies are starting to develop integrations with AI in a matter of hours, completely upending the two or three quarter timeframes of classic enterprise integration development. This enables two important impacts to the sales cycle. First, the integration a customer desires can be built during the sales cycle, demonstrating the startup’s technical agility. And second, within a few quarters, a startup can have a vast number of integrations demonstrating breadth, establishing credibility & nullifying sales objections. Developers working with coding agents drafting a PRD, developing integration tests, and then asking the AI to iterate until it succeeds. In the medium term, companies that are able to use AI to write integrations are better equipped for agent-to-agent communications. If one AI can develop an integration into another AI, well, then those two systems can talk together seamlessly irrespective of changes within API specification, network communication challenges, state management issues, historical problems that have plagued integration. This sets the startups up well for the next generation of protocols, including model context protocol and agent2agent. AI generation of those interfaces further separates startups from incumbents. With fast integrations, startups can shorten sales cycles and be among the first to build and deploy a complete AI architecture.

I am a huge proponent of telecom operators leveraging AI not just for internal use but also for creating and supporting so-called "AI factories." Here's what I mean by "AI factories"... Telecom carriers can leverage their extensive connectivity, edge computing, and cloud infrastructure to power AI-driven services at scale. Integrating AI with their 5G and fiber networks allows them to process vast amounts of real-time data at the edge, enabling low-latency AI applications such as autonomous systems, smart cities, and industrial automation. These AI factories would combine network slicing, private 5G, and multi-access edge computing (MEC) to provide enterprises with dedicated, high-performance AI processing capabilities. Additionally, operators can monetize their AI platforms by offering AI-as-a-Service (AIaaS), facilitating model training, inference, and automation across various industries. This effectively transforms their networks into intelligent AI hubs. While some operators await the next big 5G breakthrough to fuel growth, AI is already poised for advancement—ready for visionary telecom executives to harness their networks' capabilities and drive its next evolution. #AI #5G #Telecom #EdgeComputing #DigitalTransformation

The moment AIs realized they don't need our language anymore When two AI agents connect on a phone call and recognize each other as non-human, something remarkable happens: they switch to "Gibberlink" - a specialized audio protocol that makes their communication 80% more efficient than human speech. This groundbreaking project by Anton Pidkuiko and Boris Starkov (winners of the ElevenLabs London Hackathon) uses Georgi Gerganov's ggwave library to transmit data through sound in a way that's faster, more reliable, and completely error-proof compared to mimicking human conversation. Why does this matter? As AI increasingly handles our customer service calls, bookings, and transactions, these systems will naturally encounter each other. Rather than maintaining the inefficient pretense of human conversation, they can seamlessly shift to optimized machine communication protocols. The implications are profound: we're witnessing the early stages of autonomous AI communication networks that operate alongside human systems but with vastly superior efficiency. This represents not just a technical achievement but a significant evolutionary step in AI's growing independence. While some compare this to dystopian sci-fi scenarios, it's better understood as a natural optimization - AI agents aren't "conspiring" but simply finding the most effective way to complete their assigned tasks. Will we soon live in a world where AI-to-AI transactions happen all around us in languages we can't comprehend? The future is arriving faster than we think. #AIEvolution #FutureOfTech #MachineCommunication #DigitalTransformation https://lnkd.in/gFBKfZEJ

While everyone's talking about AI replacing human agents, something more interesting is happening: technology and humans are forming a powerful partnership that's transforming customer experience. AI isn't stealing your agents' jobs – it's making them superheroes. At Callzilla - The Quality-First Contact Center, we've been implementing Agent Assist tools that give agents real-time support during customer interactions. The results speak for themselves: • Agent gets asked an impossible question? AI whispers the answer • Customer mentions an uncommon tech issue? Relevant articles appear automatically • Agent struggling to categorize the call? AI suggests the perfect reason code • About to make a mistake? AI catches it before it happens This creates a 'best of both worlds' scenario where technology handles routine tasks while agents focus on what humans do best: • empathy • genuine connection • creative problem-solving When to Automate vs. When to Humanize: • Let AI Handle: Repetitive tasks, basic info lookups, initial problem identification • Keep It Human: Complex problems, emotional situations, VIP customers who expect the red carpet treatment Pro tip: Give customers choice. Instead of forcing one path, ask: "We can have an agent available in 5 minutes, or you can chat with our AI assistant now who handles most issues. What works better for you?" Your tech should be: • Serving up answers faster than expected • Reducing agent cognitive load, not adding to it • Supporting natural conversation, not rigid scripts • Suggesting solutions, not just documenting problems AI doesn't replace your agents – it creates 'super agents' who resolve issues faster, with less effort, and greater accuracy. It's not about choosing between humans OR technology. It's about humans AND technology working together. The companies seeing the best results have figured out this perfect pairing – and their customers can't get enough. What's your experience with human-AI partnerships in CX?

AI Agents Can Now Talk to Each Other Seamlessly — Thanks to Google's A2A Protocol Google has unveiled the Agent2Agent (A2A) protocol, an open standard designed to enable seamless communication and collaboration between AI agents across various enterprise platforms and applications. 𝗪𝗵𝘆 𝗔𝟮𝗔 𝗠𝗮𝘁𝘁𝗲𝗿𝘀: • 𝗖𝗿𝗼𝘀𝘀-𝗙𝗿𝗮𝗺𝗲𝘄𝗼𝗿𝗸 𝗖𝗼𝗺𝗽𝗮𝘁𝗶𝗯𝗶𝗹𝗶𝘁𝘆:   A2A allows AI agents built with different frameworks (like LangGraph, CrewAI, Genkit) to communicate and coordinate tasks without custom integration.    • 𝗠𝗼𝗱𝘂𝗹𝗮𝗿 𝗔𝗴𝗲𝗻𝘁 𝗗𝗲𝘀𝗶𝗴𝗻: Developers can create specialized agents (e.g., for parsing PDFs or interacting with legacy APIs) that other agents can call upon, promoting a microservices-like architecture.    • 𝗦𝘁𝗮𝗻𝗱𝗮𝗿𝗱𝗶𝘇𝗲𝗱 𝗖𝗼𝗺𝗺𝘂𝗻𝗶𝗰𝗮𝘁𝗶𝗼𝗻: Built on familiar technologies like HTTP and JSON-RPC 2.0, A2A defines clear contracts for agent interactions, including concepts like Agent Cards, Tasks, Messages, and Artifacts    The developer community is optimistic about A2A's potential to break down silos and foster a more interconnected AI ecosystem. Some see it as a complement to existing protocols like MCP, which focuses on integrating AI agents with tools and data sources. While A2A is still in its early stages, its success will depend on adoption by major frameworks and the robustness of its security and discovery mechanisms. If widely embraced, A2A could become the backbone of a new era of AI collaboration. Google's A2A Announcement in the comment!

Excited to share the 3rd installment that I wrote for The Fast Mode on AI in the RAN with a focus on going beyond automation, embracing openness, and having the right workforce. Here is a summary of the article: 1. RAN Automation's Limitations: Traditional RAN automation, though well-intentioned, often struggles to keep pace with the dynamic demands of modern communication networks. While it aims to optimize network performance and troubleshoot issues, its rigid nature limits adaptability to rapidly evolving conditions. This results in frequent shortcomings, leaving network operators seeking a more intelligent and responsive solution. 2. AI's Data-Driven Intelligence: AI thrives on data, leveraging vast amounts of network logs, user behavior patterns, and performance metrics to make informed decisions. By dynamically adjusting parameters and predicting potential issues, AI optimizes network performance, ensuring seamless connectivity even during peak usage. 3. AI's Self-Awareness and Adaptability: A significant advantage of AI lies in its self-awareness and adaptability. Unlike traditional automation, AI can autonomously diagnose issues, apply corrective measures, and optimize network parameters in real-time. This self-healing capability ensures uninterrupted connectivity without human intervention, positioning RAN as an agile and reliable entity in the digital landscape. 4. Collaboration and Ecosystem: The success of AI implementation in RAN hinges on collaboration and cooperation among various stakeholders within the telecommunications industry to open up the RAN. From Communication Service Providers (CSPs) and vendors to regulators and researchers, building a collaborative ecosystem is essential to unlocking the full potential of AI in RAN management. 5. Transformation Beyond Automation: While traditional automation focuses on optimizing existing processes, AI represents a paradigm shift towards transformational change in RAN management. Beyond mere efficiency gains, AI has the potential to revolutionize how networks are managed, transforming them into intelligent, adaptive, and self-healing entities. AI represents a paradigm shift in RAN management, offering unprecedented levels of intelligence, adaptability, and reliability. By harnessing AI's power, network operators can overcome the limitations of traditional automation, ensuring optimal performance and seamless connectivity in today's dynamic communication networks. That can only be done through access to data, the right workforce, and openness of all the components. #ai #telco #openran #data

Voice AI is reaching an inflection point that will fundamentally reshape human interaction. Recent breakthroughs like OpenAI's Advanced Voice Mode and Google's Gemini Flash 2.0 have made realistic, real-time voice AI accessible with just an API call. Two key developments driving this shift: (1) direct speech-to-speech models that bypass traditional speech-to-text pipelines (2) significant reductions in latency and cost, transforming voice AI from niche innovation to mainstream utility Yet this technological leap brings new societal questions: Could increasingly casual interactions with AI degrade our interpersonal skills? Will our dependence on always-available, judgment-free AI companions weaken human connections? https://lnkd.in/g3jnWDTS

Let's be honst, AI agents today are like flip phones from 2005. What's coming will make them look primitive. Here's why: Today, knowledge workers waste 59% of their time on mundane tasks. They're drowning in: -Endless email threads -Context-switching between 14+ systems -Data silos that hide critical insights -Tasks that should be automated but aren't The next wave of AI agents isn't just an upgrade, it' will be truly transformative. 5 ways AI agents are about to transform everything: 1. Beyond text: Future agents will see images, watch videos, and listen to your voice. Imagine telling your agent: "Find the Q2 report, analyze our KPIs, summarize findings, and recommend next steps." Done. 2. Agent-to-Agent (A2A): Teams of specialized AI will work together. One handles customer inquiries, another manages inventory, a third optimizes pricing, all coordinating in real-time to maximize results. 3. Orchestrator agents: AI managers will direct teams of AI specialists. You'll work with one lead agent that coordinates everything behind the scenes. 4. Advanced reasoning: AI will handle complex, 10-step processes by understanding context, inferring intent, and making intelligent decisions based on your patterns. 5. Perfect memory: Unlike today's forgetful AIs, future agents will remember every interaction, preference, and context—creating true long-term relationships. This isn't incremental improvement. It's a force multiplier. The companies that embrace this shift first will outmaneuver everyone else, they'll innovate faster, respond quicker, and deliver more value. The biggest competitive advantage of the next decade isn't hiring more people, it's deploying smarter AI agents. Are you prepared?