AI AGENTS FOR
ForestHub is the Edge AI and Agents Orchestration Platform where workflows are graphs and the AI is one node among many — not the program itself. Built for industrial control where every decision must be inspectable, replayable, and bounded.
AI AGENTS FOR
The Edge AI and Agents
Orchestration Platform. How it works
ForestHub is the Edge AI and Agents Orchestration Platform where workflows are graphs and the AI is one node among many — not the program itself. Built for industrial control where every decision must be inspectable, replayable, and bounded.
ORCHESTRATE YOUR ENTIRE DEVELOPMENT PROCESS.
From visual workflow to running engine on your Linux edge device — ForestHub covers the full lifecycle: design, deploy, run, observe.
Workflow Builder
Visual canvas for industrial agent workflows.
Author the sense-reason-act loop as a graph on a canvas. Wire sensors, deterministic operations, LLM agents, and actuators into a single workflow — then deploy as a versioned artifact to any registered device.
Explore the Platform →THE RIGHT INTELLIGENCE AT THE RIGHT LAYER.
Not every task needs a 70-billion-parameter model. ForestHub orchestrates exactly what each task requires.
An industrial workflow is rarely an end-to-end pipeline of the most expensive cloud model. It's a cascade: hard rules where safety matters, classical ML where proven classification is enough, small language models on-device where you need language but the data stays local, and large cloud models only where the extra quality earns the data egress.
On each node, you pick the tier that fits the task. Inference cost drops where it can, data stays where you want it, latency stays in the millisecond-to-second range, and the system no longer depends on a single cloud LLM vendor.
- 01
Rule-based Logic
On-device · Sub-millisecond · Data: stays local
Deterministic if-then logic, thresholds, state machines. Not AI, fully predictable, sub-millisecond. For safety-relevant decisions — "under 80 °C: do nothing, above 90 °C: shut down" — this remains the right tool, because it's auditable without training data.
- 02
Classical Machine Learning
On-device · Milliseconds · Data: stays local
XGBoost, random forests, compact neural nets for anomaly detection and classification. Runs on-device in a few megabytes of memory, millisecond inference. Proven models for the most common industrial tasks — vibration-based machine monitoring, line-side image recognition, predictive maintenance — none of which need a language model.
- 03Sweet spot
Small Language Models (SLMs)
On-device · Seconds · Data: stays local
Models like Phi, Gemma, or Llama in the 1–3B parameter class. Architecturally complete language models, just smaller. They run on-device on industrial PCs or gateways in a few gigabytes of memory, with response times in seconds. Enough for classification, short diagnostic dialogs with the technician, and document RAG — entirely on your hardware.
- 04
Mid-sized Open-Weight Language Models
On-premise GPU · Seconds · Data: stays local
70B-class models such as Llama or Qwen, runnable on NVIDIA Jetson clusters or a dedicated on-premise GPU server. Business value: demanding reasoning on-premise, without data flowing into someone else's cloud. The tier for high-stakes diagnostic work under strict data-protection requirements.
- 05
Frontier Language Models via Cloud
Cloud · Seconds · Data: leaves your network
Claude, GPT, Gemini. Highest quality, best multilingual performance, full provider toolchains. Reserved for the few cases where the extra quality justifies sending data outside — complex customer-facing dialog or multilingual reasoning a 3B model can't carry.
For roughly nine out of ten industrial tasks, a small language model on your own hardware is enough. Frontier models remain the exception, used deliberately — not the default. That is the honest answer to "What does this cost us?" and "Where does our data go?".
The graph is the program.The AI is a node.
Most AI agent frameworks let the LLM run the show — it loops, calls tools, decides when it's done. For consumer assistants that works. For industrial control, where a wrong decision means a relay actuates or a setpoint changes, it doesn't. ForestHub flips it: you draw the workflow as a graph, the LLM is one node among many, and every possible path is visible at design time. And a node need not be a frontier LLM: the graph orchestrates a cascade of rule-based logic, classical machine learning, on-device small language models, and large cloud models — pick the tier that fits each node. Inference cost stays low where a small on-device model is enough, and data leaves your network only when the extra quality earns it.
RIGHT WHERE IT MATTERS — AI THAT WORKS.
Real-world Edge AI applications — from predictive maintenance to autonomous sensing. See how ForestHub turns embedded hardware into intelligent systems.
AI-Powered Diagnostics Right at the Machine
On-device agents for industrial diagnostics: ingest sensor data, analyze locally, identify root causes, deliver actionable instructions. No cloud round-trip, no latency.
Buildings That Think — Locally and in Real Time
Building technology OEMs and integrators use ForestHub to build AI-driven automation systems that optimize heating, ventilation, lighting, and energy consumption autonomously and locally — no cloud required, full data control.
Predict Failures Before They Happen
Machine builders and plant operators use ForestHub to deploy AI models directly on their equipment — detecting wear, anomalies, and imminent failures early, without sending sensor data to the cloud.
Intelligent Sensors That Decide on Their Own
Sensor manufacturers and system integrators use ForestHub to equip their sensors with embedded AI — for classification, filtering, and decision-making right at the sensor, without transmitting data first.
BUILT FOR EU COMPLIANCE.
European regulations demand more from AI systems than ever before. ForestHub meets them by architecture — not by afterthought.
Data Sovereignty
Data sovereignty is a consequence of the engine's deployment model — workflows run on hardware you control, no telemetry to a vendor cloud unless you wire it. Where cloud-based AI platforms force you to patch around regulatory requirements, the graph-first, Linux-edge architecture solves them structurally.
GDPR / DSGVO
All processing on-device. No personal or operational data leaves your hardware. GDPR compliance by design — not by policy.
EU AI Act
Risk classification, documentation, and human oversight built in from day one. Full traceability for high-risk AI system obligations.
Cyber Resilience Act
Secure-by-design updates, SBOM support, and vulnerability management. CRA-ready for connected products out of the box.
Open-Core Model
Auditable, transparent, open-source core. Inspect every model, every decision path, every deployment. No black-box AI — full visibility into what runs on your hardware.
Auditability & Traceability
Auditability is a consequence of the graph: every LLM decision is a wire on a canvas, every run is a structured event log, CI replay is a primitive. Audit-ready for EU AI Act compliance and enterprise governance.
EXPERIENCE THAT CREATES IMPACT.
After more than a decade in the embedded world, we've seen how much potential lies in data that's never used. Technical knowledge exists - in manuals, in minds, in systems. But it's not where it's needed: with the technician at the machine, with the customer with a question, with the developer with a problem. ForestHub builds the bridge. We make knowledge accessible, intelligent, and actionable - exactly where it creates value and not in the cloud.
Meet the team
READY TO BUILD EDGE AI?
The ForestHub Orchestration Platform is live. Register free and deploy your first workflow to a Linux edge device — graph-first, audit-grade, without cloud dependency.
Free — no credit card required.
NEWS FROM FORESTHUB.
The latest updates from our team.
Embedded AI as a Service Technician
Costly downtime and service tickets are among the biggest cost drivers in manufacturing. Together with our partner Grossenbacher Systeme, we show how embedded AI — combining small and large language models with agentic orchestration — makes a “virtual service technician” possible, running entirely locally on the edge controller.
May 12, 2026The ForestHub Orchestration Platform Beta Is Live
Starting today, engineering teams can register for free at app.foresthub.ai. Visual workflow builder, Go-based engine in Docker, multi-LLM routing, industrial protocols first-party — inspectable, replayable, auditable, bounded by design.
We're Open-Sourcing boardsmith
Contact
DISCUSS PROJECT.
Sales & Partnerships
Our team supports you in evaluating ForestHub solutions for your organization. From initial idea to productive deployment.
root@foresthub.ai
Headquarters
ForestHub GmbH
Schluchseestraße 25
78054 Villingen-Schwenningen
Germany
Let's Talk About Your Project
Whether you're exploring Edge AI for the first time or ready to deploy — we're here to help. Select a topic and tell us about your requirements.