Cisco expanded Webex to add AI agents that can participate in meetings and calls, and to connect customer-service work to continue after interactions end. Most features and products are planned for early 2027, with Dialog beta targeted for the first quarter of 2027. Webex gets a redesigned, personalized meeting view plus new control, security, and integration layers (including planned Claude and Dots integrations) and more on-prem AI and workplace hardware/software updates.
Meta’s Muse assistant launched as a dedicated app for iPad, following its mobile debut. Muse reached 6.6 million iPhone and Android installs after launching on September 8. The iPad release adds new “connectors” (including several small-business tools) and expands the set of retailers it can help users shop with on Meta’s behalf.
TheCUBE’s coverage of ‘The AI ROI in Contact Center Summit’ described how contact centers are shifting from measuring call speed toward blending agentic AI with humans to resolve customer issues and improve trust. Only 66% of actual users said contact centers have gotten better with AI, compared with 99% of business practitioners. Companies are therefore moving to “conversation to completion” metrics and hybrid operating models that preserve context across handoffs and assign AI to high-volume, low-stakes work while humans handle complex cases.
Semwright launched today as an open-source Rust runtime for AI agents that can use structured drivers to interact with desktop and professional software and related project data.
It supports MCP while not being tied to a single protocol, model, or agent.
It changes how agents are built by centralizing permissions, execution, dependencies, artifact handoffs, and verification so developers can reuse integration infrastructure across multiple applications rather than rebuilding it each time.
Infor is rolling out AI-powered process automation that embeds industry-specific agents and workflows into its applications, pairing that with implementation expertise from forward-deployed engineers. It says the prototype-to-production timeline is now about one to two weeks, or at worst three weeks. As a result, customers move faster to production while getting observability and auditability for automated decisions.
Qlik built Qlik Answers, a grounded, source-cited enterprise AI assistant on Amazon Bedrock, and has expanded its use since its February 2026 general availability. Since launch, Qlik’s Discovery Agent has surfaced more than 100,000 discoveries, and most Qlik Cloud accounts with agentic tools enabled are actively using them. The system’s architecture now separates fast routing from deeper reasoning and uses Bedrock Guardrails plus grounding-validation to deploy safely and at regional scale.
AWS DevOps Agent triages an incident and generates an investigation summary, which then triggers an automated remediation workflow built with EventBridge, Lambda Durable Functions, and Bedrock to propose and (after approval) apply fixes.
A six-week structured program helped business professionals who use AI daily but do not code build working multi-agent AI prototypes instead of only discussing them.
Cornerstone OnDemand built Orion AI, a multi-agent system using Amazon Bedrock and Strands Agents to automate database diagnosis and operational workflows that previously required manual handoffs. Database diagnosis time fell from 45 minutes to 10 minutes, a 78% reduction. The change turned reactive, multi-step firefighting into faster, guided investigations with single-interaction lifecycle handling and reduced alert noise.
Meta and Sierra released the Personal Agent Protocol (PAP) to define how personal AI agents authenticate and act across merchants’ services. Version 0.1 is scheduled later this month. As a result, participating businesses can set access to public data, read-only account details, or write actions like returns, with payments and push notifications deferred to later extensions.
OpenAI’s product lead Alexander Embiricos is scheduled to appear at TechCrunch Disrupt 2026 in San Francisco to discuss OpenAI’s next-step personal AI agent, Dots, shortly after its launch. The conference runs October 13–15. As a result, Embiricos will hold a session focused on what comes after the chatbot and the privacy and trust questions tied to more autonomous AI agents.
TechCrunch Disrupt 2026 will host interactive roundtables during October 13–15 in San Francisco that cover topics including AI, fundraising, hiring, enterprise software, and physical AI. The lineup includes 10,000+ founders, investors, operators, and tech leaders attending Moscone West for three days. Attendees can save $100 on their pass (and get 50% off a second pass) to join the sessions and ask questions with founders and experts.
Microsoft demonstrated an updated Copilot AI system at its Windows and Surface event that can access local files and take OS-wide actions. The demo had an Autopilot request to help with filing taxes by handling an email from an accountant, using a Hybrid Intelligence setup that mixes local and cloud AI models. This expands Copilot from assisting in apps to operating across Windows and your documents more directly.
Kore.ai launched Autoloop, an optimization engine that continuously tunes enterprise AI agents after they go live on its Agent Platform. Autoloop cites its 2026 Agent Productivity Index showing 79% of enterprises had reversed an action taken by an AI agent. This adds automated, goal-based iteration using trace-based validation and targeted blueprint-based fixes instead of manual one-off patching.
Opposable launched a browser automation service that lets Claude Code, Codex, or any MCP agent interact with real iPhones or Android phones in apps you’re already signed into. The service launched today. As a result, some tasks can run on the phone without a model, and iPhones can be controlled via Bluetooth or a small USB stick.
Clippo launched as a Windows app that lets you orchestrate AI agent teams on an infinite visual canvas instead of using terminal tabs. It provides parallel coding by linking nodes and positions on an infinite canvas, with zero cloud and no logins. As a result, developers can manage multiple AI coding agents visually and run them on device to reduce token use.
Stuut raised a $52.5M Series B led by Insight Partners to deploy AI agents that track invoice chains across enterprise order-to-cash workflows. The round closed in 24 hours, following a $29.5M Series A 10 months earlier and lifting total funding to $93M. Stuut says the software shifts collections and payment matching from staff effort to automated workflows and expands it to more enterprises.
SquadRoom launched with AI workflow automation that assigns tasks to a lead agent and specialist agents, then shows a step-by-step report with sources after you return. It starts with 21 followers on the page. Users can bring their own AI key and can run a live demo without signing up, with items requiring approval held back until you OK.
The Sequence Learning Loop (Issue 946) discussed how OpenAI’s September 29 DevDay releases and Google’s September 30 Gemini 4 Argon focus on the economics of running long, multi-step AI tasks. On September 29, it highlighted DevDay changes aimed at sustained execution costs and infrastructure. It argues that competition is shifting from model intelligence alone toward the full completed workflow, including tooling and budget for repository edits, test runs, and reviewable outputs.
The keynote speaker at LDX3 described how AI agents are reshaping software development practices and platform activity, based on visits to OpenAI and Anthropic and data from GitHub, Factory AI, and Linear. GitHub reported a ninefold increase in agent-authored pull requests over eight months. As a result, engineers increasingly rely on multiple parallel agents and non-IDE agentic workflows, while companies build internal agent harnesses and agent-generated issues and PRs begin to outnumber human ones.
Circeus argues that business AI adoption is slow because AI is often added as a separate assistant instead of being embedded into the software systems that already run everyday workflows. It says meaningful workflow implementation in the everyday economy is closer to 4–5%, about one in twenty. The company’s approach is to acquire mission-critical software businesses and evolve their products into AI-native systems with bounded actions, human approval, and ongoing measurement of task completion and errors.
Google announced Playground, a browser-based platform for making custom games from AI prompts. It is available starting today in the US for users 18 and older at playground.google. Users can create and test games by typing prompts, which removes the need for coding experience.
Zernio is marketing infrastructure that provides a single API for integrating social, ads, messaging, and calls so teams can launch without building separate channel integrations. It supports phone numbers and messaging across 57 countries. As a result, product teams can connect accounts and run campaigns through Zernio’s API or its hosted MCP server instead of managing per-channel authentication and approvals themselves.
Leanback launched its second release, presenting itself as a personal AI assistant for engineering teams. It promises to produce an action plan in Slack using inputs like tickets and calendars after checking in with the team. As a result, teams can manage engineering work from Slack with the assistant providing the planned next steps.
BotBus connects local coding agents to your phone so you can manage tools like Codex, Claude Code, Hermes, OpenCode, and DeepSeek across computers. It launches today. It adds mobile access (iPhone, Apple Watch, Android) plus macOS screen control and shared web previews while letting supported tasks keep running as approvals and file viewing are handled remotely.
Every AI story that matters,
in your inbox by 8am.
TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the
day in two minutes. Follow companies and topics for alerts, or get the
briefing in Slack. Free, no spam, unsubscribe anytime.
Reading TLDRocket needs no cookies, and the readership counts we rely on come from
our own cookieless analytics. Google Analytics is the exception: it sets cookies and
reports to Google, so it stays switched off until you allow it. You can change your
mind any time from “Cookie settings” in the footer.
Strictly necessary
Session security and form protection (tldrocket-session,
XSRF-TOKEN, 2 hours). The site cannot work without them,
so they need no consent.
Always on
Google Analytics 4 (_ga,
_ga_<id>, up to 2 years). Measures which
stories and sections readers use. Google acts as a third-party processor and may
store the data outside the EU. No advertising, no profiling, no data sold.
