{{first_name | Hey}}, welcome back.
In today’s issue we talk about the new voice agents by ElevenLabs - fastest in its class, how Anthropic is using its AI models to uncover open-source security flaws, and OpenAI Academy’s Livestream helping teams measure their AI investments.
These are today's updates.
ElevenLabs launches Eleven Realtime for voice agents
Anthropic launches a free AI scanner for open-source vulnerabilities
OpenAI’s masterclass on how to know if your AI investment is paying off
Tools, jobs, resources, and last issue’s winning prompt ⬇️
Spotlight: OpenAI
Image Source: OpenAI’s ‘Impact Math: Measuring the ROI of LLMs’ page
You’ve rolled out AI across your team, but how do you know it is delivering meaningful value? OpenAI Academy’s Impact Math: Measuring the ROI of LLMs explores how to measure the impact of AI beyond usage numbers.
Register for the livestream if you want to learn:
How to measure the business impact of LLMs.
Which metrics help make the case for AI investment.
How to connect AI adoption to measurable business outcomes.
It’s happening on Oct 28 – Oct 29, 2026
Top News
ElevenLabs launched Eleven Realtime, a speech-to-speech model for voice agents, in two versions. v1 Max responds in 0.57 seconds and has a score of 98.8% on a spoken reasoning test. ElevenLabs says no model on Artificial Analysis’s leaderboard is both faster and more accurate.
Max is great for support calls and lead qualification while Mini suits assistants and tutors. Both cost $0.08 a minute plus token charges, in ElevenAgents or an OpenAI-compatible API. You can’t swap the underlying models. It’s just Max that offers Zero Retention and HIPAA settings, so check that before taking customer calls. You can try Eleven Realtime here.
Quick question: Would you prefer to read newsletters, like Neatprompts, yourself or delegate it to an AI agent?
Anthropic launched OSS Scanner, a free service where its strongest models scan important open-source projects for security flaws. It also emails the maintainers a report with a suggested patch. Opensource maintainers can enroll through Anthropic’s OSS Scanner page with a pull request on GitHub.
Anthropic says its first round found 77 critical and 803 high-severity issues. One was a use-after-free bug in curl’s HTTP/2 handling, fixed in version 8.23.0. Reports are sent to maintainers before anyone at Anthropic checks them, so some may be wrong. It seems open-source tools your teams use may get more security fixes. Keep installing the updates promptly. To know more read the announcement.
Signals
Tools
Notion Auto: Notion’s rebuilt Agents run for hours and cost far less
Ilha: Shows how much of your Claude Code, Codex and Grok Build limits you've used
Intelli for ChatGPT & Claude: Run WhatsApp, Instagram & web chat from ChatGPT or Claude
Resources
Poll
Choose a prompt you want us to publish next
Market
Funding
Atomic Machines raised $250M to build micro machines directly from code
ASI raised $325M to expand AI systems for critical physical infrastructure
Roles in AI
Vulnerability Analyst at xAI
Engineering Manager, Inference at Mistral AI
Tutorial
Build your own team of named Bots that can work inside the apps and websites you already use. This beginner-friendly video shows how to set up Grok Bot, use the Cloud Computer, connect multiple Bots in one thread, and build nine practical bots for work and everyday life.
Prompt of the Day
AI Use Case Discovery Assistant
When to use this?
Use this prompt to find practical AI opportunities in your team or organization, prioritize the ones worth pursuing, and turn them into actionable next steps.
**Role:** Act as an enterprise AI strategist who helps business leaders identify where AI can deliver measurable value. Focus on real business problems, practical implementation, and opportunities that fit the organization’s goals and constraints.
**Context:**
- Organization or team: [describe your business or function]
- Current priorities: [e.g., reduce costs, improve productivity, increase revenue]
- Biggest operational challenges: [list recurring bottlenecks or...
{Click on the button to view full prompt}Get more such prompts in the Prompting Playbook (free for you)
Stay curious, {{first_name | reader}}
PS. If you missed yesterday’s issue, you can find it here.
