Tuhüt is an autonomous gateway agent that lives directly on your test bench. Flash firmware, capture logs, and sync evidence asynchronously—cutting your AWS bills by up to 90%.
Traditional HIL platforms force you to keep expensive AWS EC2 instances running 24/7 just to hold a serial connection open. Tuhüt flips the model: smart edge agents do the heavy lifting, the cloud just orchestrates.
The lightweight Python agent runs on a local PC or Raspberry Pi, managing the USB connection to your test device.
When a test is triggered via API, the agent flashes the device, captures CAN/Serial data, and runs standard Pytests.
Results, artifacts, and logs are bundled and pushed to cloud storage asynchronously. You only pay for storage, not compute.
Traditional CI runners on AWS EC2 can cost thousands per month just waiting for hardware responses. See how moving to edge computing impacts your bottom line.
Testing embedded hardware is inherently messy. We built Tuhüt to handle the chaos locally so your cloud CI remains clean, reliable, and deterministic.
Native integration with NXP LinkServer CLI. Automatically drops hex/bin files onto target boards and validates successful boot sequences.
Plug in a board and the agent auto-detects ttyACM ports, matching them to test profiles without manual config files.
Captures raw serial logs, CAN bus dumps (candump), and pytest XML results into a single immutable artifact bundle per run.
If the cloud connection drops mid-test, the local agent continues running. Evidence is buffered and synced when connectivity restores.
Hardware is physically isolated. Test definitions are pulled down, executed in a clean environment, and wiped post-run.
Triggers testing via secure webhooks from GitHub Actions, GitLab CI, or Jenkins. No reverse-tunneling or complex VPNs required.
"Before Tuhüt, we were managing a complex VPN setup just to trigger tests on our lab bench from GitHub Actions. The network instability caused 40% of our test failures. Moving to an asynchronous edge agent solved our flakiness overnight."
Join the early access program. We are currently onboarding embedded engineering teams with existing Pytest automation suites.