TrueFoundry + Seldon: Unified Control Plane for Enterprise AI

Auf Geschwindigkeit ausgelegt: ~ 10 ms Latenz, auch unter Last
Unglaublich schnelle Methode zum Erstellen, Verfolgen und Bereitstellen Ihrer Modelle!
- Verarbeitet mehr als 350 RPS auf nur 1 vCPU — kein Tuning erforderlich
- Produktionsbereit mit vollem Unternehmenssupport
Today we're sharing that TrueFoundry has acquired Seldon's business and customers.
For over a decade, Seldon has been the backbone for real-time ML inference at some of the most demanding enterprises in the world — across banking, retail, telecom, insurance, and healthcare. They went deep on one of the hardest problems in production AI: serving classic ML models in real time, at scale, on Kubernetes. Enterprises like Best Buy, Verizon, and PayPal trust them to run it.
Seldon focused on perfecting that. As the market expanded into autonomous agents, that's the layer their customers now need to add — and that's where TrueFoundry comes in.
TrueFoundry is a Control Plane for Enterprise AI, running on Kubernetes, that lets companies launch AI at scale with the right governance, observability, and cost controls. Our AI Deploy product runs and scales Agentic AI and ML workloads. Our AI Gateway is the control plane for Agentic AI — connecting, observing, and governing every LLM, agent, and tool call from one place. This year we launched our Agent Gateway, extending that control plane deeper into how agents operate.
That's the bridge we're giving Seldon's customers: keep the production ML they've built, and gain a clear path into Agentic AI on the same Kubernetes foundation they already run.
This is the second major step we've taken this year, after launching the Agent Gateway. Here's why we did it, why now, and why Seldon.
Every enterprise now lives on both sides of AI
Talk to almost any enterprise AI team today and you'll hear the same thing. They're running traditional ML models — fraud detection, churn prediction, recommendations — and they're building agentic workflows that reason, call tools, and act on their own.
These used to evolve at different speeds. They don't anymore. Both are becoming mission-critical at the same time.
Managed as two separate problems — two vendors, two governance models, two sets of tooling — that's expensive and fragile. For banks, hospitals, insurers, and retailers, a model degradation or a governance blind spot becomes a customer-facing problem in minutes.
These enterprises don't want to stitch together two stacks. They want one platform, from one vendor, that runs on the infrastructure they already have.
Why Seldon, and why now
Seldon's customers are exactly the enterprises living on both sides of that line. They've built world-class production ML — and the agentic half is the natural next step for them.
Why now: agents crossed from experiment to production this year. The enterprises running Seldon's ML are the same ones being asked to ship agentic applications now — and they need a path that builds on what they already trust, not a second platform to stand up from scratch.
We're the right home for that step for two reasons.
Shared foundation. Seldon and TrueFoundry were built on the same conviction: portable, Kubernetes-native infrastructure that runs wherever the enterprise already operates. No new philosophy to adopt, no platform to re-architect around. Their teams are already where ours lives.
A real path into Agentic AI. This is what TrueFoundry does well, and has perfected over years working with enterprise customers. Through our AI Gateway, Seldon's customers can connect, observe, and govern agents alongside the ML they already run — turning a decade of production ML investment into a launchpad for agentic applications, not a thing to leave behind.
What this means for Seldon's customers
If you're a Seldon customer, the first thing to know is this: your ML keeps running on our AI Deploy platform.
You keep the Kubernetes infrastructure you've invested years in. We're actively working with you to build the path forward — continuity of support today as you adopt the TrueFoundry platform for your AI workloads, and a bridge into Agentic AI through our AI Gateway that brings it all under a single control plane for enterprise AI. That work is underway now, and we'll be in touch directly.
What you gain is a single platform for what's next: your production ML, and the agents, LLMs, and tools you'll build around it — on your terms and your timeline.
What this means for TrueFoundry's customers
TrueFoundry's customers include some of the biggest enterprises in the world — Mastercard, Cargill, Cox Communications, Synopsys, ResMed, Siemens Healthineers, Pfizer, and Automation Anywhere — running business-critical AI through the Gateway today. Together they process over a trillion tokens a day. Seldon's customers sit in exactly the same category, and the verticals Seldon matured in help us extend the enterprise coverage we already have.
For our existing customers, nothing changes with this acquisition — and they gain something. The top enterprises joining us from Seldon bring deep, vertical-specific expertise that feeds directly into our roadmap. As they become TrueFoundry customers, that experience makes the platform better for everyone, and moves us further along our path to being the Control Plane for Enterprise AI.
The control plane for enterprise AI
When we started TrueFoundry, we believed enterprises would need one place to build, ship, and govern their AI — not a pile of disconnected tools. As AI moved from predictive models to autonomous agents, that need only got sharper. Our vision has always been to make TrueFoundry the Control Plane for Enterprise AI.
Acquiring Seldon's business lets us extend that vision to their customers and moves all of us closer to it. It won't be the last step we take. The market is moving fast, and the enterprises we work with don't have the luxury of waiting for it to slow down. Neither do we.
Seldon customers: we'll be reaching out directly, but if you have questions now, talk to our team.
Evaluating TrueFoundry? Here's how this fits into where we're headed — schedule a call.
TrueFoundry AI Gateway bietet eine Latenz von ~3—4 ms, verarbeitet mehr als 350 RPS auf einer vCPU, skaliert problemlos horizontal und ist produktionsbereit, während LiteLM unter einer hohen Latenz leidet, mit moderaten RPS zu kämpfen hat, keine integrierte Skalierung hat und sich am besten für leichte Workloads oder Prototyp-Workloads eignet.
Der schnellste Weg, deine KI zu entwickeln, zu steuern und zu skalieren














.webp)

.webp)





.webp)
.webp)
.webp)
.webp)
.webp)



