At Open Source Summit in Minneapolis, I was walking the show floor when I came across the Edera booth and stopped to learn more. The company is focused on securing workloads once they reach production by isolating them inside microVMs. The approach piqued my interest, so I sat down with Chief Marketing Officer Kaylin Trychon to learn more.
Edera is a two-year-old startup founded by former Chainguard employees. It describes itself as a “production-grade sandbox” for running untrusted workloads, including AI-written code, AI agents, and open source software. The rise of AI agents is making that need more urgent as organizations begin putting increasingly autonomous software into production.
When I asked how Edera differs from Chainguard and Minimus, Kaylin explained that Edera complements hardened-image vendors by focusing on what happens at runtime. It runs workloads inside microVMs so that if a container or AI agent is compromised or behaves unexpectedly, the problem stays contained instead of spreading. Edera has also partnered with Minimus.
Kaylin also covered Edera’s $20 million Series A led by Microsoft’s M12 and its next area of focus: GPU isolation and slicing for neocloud providers.
Check out the full interview above for Kaylin’s explanation of how Edera approaches runtime isolation, AI agents, and GPU security.
Chapters
00:00 — What Edera does
00:50 — Edera vs. Chainguard and Minimus
01:45 — Edera’s origins
02:35 — AI agents in production
03:30 — Production-grade sandboxing
03:55 — Agent adoption
05:05 — Governance
05:30 — Funding and growth
05:55 — GPU isolation
06:25 — Neoclouds and GPU utilization
07:00 — What’s next
Pau for now..
Leave a Reply