Subvolts take AI makes security cameras smart, detecting and stopping threats before they happen. Dec 16, 2025 - Developers
- Learn how to build specialized multimodal models for real-world applications like security.
- Business
- Understand how AI can enhance physical security operations and prevent incidents proactively.
This interview covers a practical application of AI in physical security, detailing a company's approach and technology.
Building the “See Something, Say Something” AI for Every Camera
Audio streams from a16z. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
AI + a16z features Martin Casado interviewing Shikhar Shrestha, CEO of Ambient. Shrestha discusses his mission to make security cameras intelligent, inspired by a childhood robbery. Ambient's AI monitors feeds in real-time to detect and prevent threats, building its own reasoning VLM called Pulsar. The future of security involves AI not just detecting but also responding to incidents.
This episode explains how AI can watch security cameras and spot trouble before it starts, like a vigilant guard. It's about making cameras intelligent so they can help keep places safe by detecting dangers and even acting on them.
This episode offers insights into applying agentic AI to physical security. Listeners will learn how Ambient uses AI to proactively monitor camera feeds, prevent incidents, and build specialized multimodal models. It's for security professionals and AI enthusiasts interested in real-world applications beyond typical chatbots. This aligns with frontier labs' push into multimodal AI and agentic systems for practical, impactful use cases.
- Hosta16z
- GuestShikhar ShresthaCEO and cofounder, Ambient
- Episode Title: Building the “See Something, Say Something” AI for Every Camera
- Show: AI + a16z
- Length: 39 minutes
- Guest's Company: Ambient
- AI Model Mentioned: Pulsar (reasoning VLM)
a16z's Martin Casado sits down with Shikhar Shrestha, CEO and cofounder of Ambient, the company bringing agentic AI to physical security.
Shikhar shares how a traumatic armed robbery at age 12—and a security camera that no one was watching—sparked his mission to make every camera intelligent.
They discuss how Ambient's AI monitors camera feeds in real-time to detect threats and prevent incidents before they happen, navigating COVID as a physical security company, building their own reasoning VLM called Pulsar, and why the future of security is AI not just detecting threats but automatically responding to them.
If you enjoyed this episode, please be sure to like, subscribe, and share with your friends.
Follow Shikhar on X: https://x.com/shikharshrestha
Follow Martin on X: x.com/martin_casado
Check out everything a16z is doing with artificial intelligence here, including articles, projects, and more podcasts.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
- agentic AI
- Shikhar Shrestha
- AI that can act autonomously to achieve goals.
- VLM
- Shikhar Shrestha
- Visual Language Model, an AI that understands both images and text.
24 minHow AI Is Rewriting the Rules of Cybersecurity | Truffle Security & Socket
WithDylan Ayrey, Feross Aboukhadijeh
47 minInside vLLM: The Engine Powering Open-Source AI
WithSimon Mo
41 minBuilding Self-Accelerating AI with Mirendil
WithBehnam Neyshabur, Harsh Mehta
43 minIdeogram’s Open-Weights Image Model and the Future of AI Design
WithMohammad Norouzi
Questions people ask
- What is Ambient's mission?
- Ambient's mission is to make every camera intelligent, inspired by a personal experience with a security camera that was not being watched.
- How does Ambient's AI work?
- Ambient's AI monitors camera feeds in real-time to detect threats and prevent incidents before they happen, and aims for automatic responses.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to a16z.