BLOG: From FIRST PROTOTYPE to Scaled Deployments →
The fastest way from idea to working physical system. Ready to scale when you are. Deploy your code–business logic, sensor drivers, or ML modules–to real hardware in minutes. Swap modules with a click. Debug live. Use any framework. make87 handles the rest.
Challenges
The everyday obstacles that make robotics and edge AI harder than it should be.
Cross-architecture builds, SSH setup, and dependency mismatches. It’s not hard — just constant friction that kills focus.
You either tether everything to your desktop or start optimizing for edge too early. Both slow down meaningful progress.
Work spreads across nodes, sensors, and models — but the shared understanding lives on a whiteboard or in someone’s head. Alignment breaks fast.
Change code, rebuild, redeploy, test, repeat. It should take two minutes — not an hour — to see if your fix even works.
When something breaks, you just want to look inside, stream logs, or patch remotely — without wiring up VPNs and scripts first.
Features
From deployment to debugging — everything that used to take hours now happens in minutes. These are the core features that make it possible.
Assign a node to an application and it just runs — no manual setup, no config files, no SSH. make87 handles builds, architecture, and deployment automatically.
Open any app in the web IDE or attach your local editor via SSH. make87 manages secure forwards automatically — no VPNs, no setup.
The system graph isn’t a mockup — it’s your live setup. Everyone sees what’s connected to what, which versions are deployed, and how interfaces link.
Shift compute between edge and cloud seamlessly. Networking, ports, and environments configure themselves — just choose where to run your application.
View logs, metrics, and statuses from every node. Open remote shells instantly to patch issues and verify performance in real time.
How It Works
Connect hardware, import code, design your system, debug remotely, and deploy anywhere — all from one dashboard.
Run one command on your edge hardware — it connects automatically. No VPNs, keys, or manual config needed.
Link your GitHub account and import existing apps. Each repo just needs a simple make87.yml to define its interfaces — dependencies stay in your Dockerfile as usual.
Visually link apps, nodes, and interfaces in the system designer — then deploy them to real hardware with a single click. make87 builds for the right architecture, versions every release, and lets you roll back instantly when needed.
Connect your favorite IDE and debug applications running on any device, anywhere — full remote introspection without extra setup or tunnels.
Build and test for free, collaborate with your team, then scale to production without rebuilding.
Ideal for individuals prototyping robots, CV systems, and sensor applications.
$0 FLAT FEE
SYSTEMS
2 active systems
COLLABORATION
Solo use only
DEPLOYMENT
Unlimited self-hosted nodes
OBSERVABILITY
Basic metrics and logs
MARKETPLACE
Access to pre-built applications
STORAGE
500MB included to get you started
SUPPORT
Community Discord & documentation
For teams iterating fast on prototypes and ready for pro features.
› Everything in Free, plus:
SYSTEMS
Up to 30 systems
COLLABORATION
Multi-user workspaces
DEPLOYMENT
Access to managed cloud nodes
CUSTOMER PORTALS
Expose UIs with managed access
STORAGE
Flexible, expandable as needed
SUPPORT & SLA
Email support
For production deployments and advanced enterprise needs.
› Everything in Teams, plus:
SYSTEMS
No system limit
MULTI-REGION ORCHESTRATION
Management nodes deployed globally
IDENTITY & ACCESS
SSO, SAML & SCM Authentication
CUSTOMER PORTALS
Expose UIs with SSO access
LICENSING
Site-license terms available
SUPPORT & SLA
Dedicated CSM (priority response)
Updates, tutorials, and case studies from the make87 team — covering edge AI, robotics, and production-grade prototyping.
Tutorial
Develop and debug robots remotely with make87’s Dev Mode — full IDE and SSH access, live sensor feedback, and real-time iteration on real hardware. Modern software workflows meet robotics.
Deep Dive
Video decoding consumes substantial CPU resources in edge AI systems. Our analysis using Intel QuickSync shows that offloading decode to integrated GPUs dramatically reduces CPU load and power consumption, enabling more camera streams per device.
