Production work should start from context
Humans and agents should not start from zero every time something changes. The system should maintain evidence, ownership, impact, and the action path around production work.
About Tero
Production needs its own AI leverage.
AI is changing how software gets written. More code ships, more systems change, and more operational complexity lands in production. Tero exists to change the production workflow from human-speed investigation to a continuous loop of evidence, issues, and safe resolution.
Why logs first
Logs are where production leaves the messy truth.
Failed connections. Bad retries. Sensitive fields. Expensive noise. Repeated warnings. Misconfigurations. Operational risk that never becomes a clean alert.
For decades, logs have mostly been used after the fact because they were too large and noisy to read continuously. That is changing.
Tero starts with logs because they are the richest underused evidence source in production. If we can turn logs into issues teams can understand and fix, we can prove a new operating model for production itself.
What we believe
Principles for production software in the AI era.
Finding issues is only half the loop
A useful system cannot stop at recommendations. Tero should help route each issue through the safest path to resolution.
New work needs new action surfaces
When the right production change mechanism does not exist, we should build it. Telemetry Policies are the first example: small, reviewable, measurable changes where telemetry flows.
Existing systems still matter
Observability platforms, collectors, runtimes, Git, tickets, and engineering workflows are not going away. Tero should work with the stack teams already run.
Open infrastructure earns trust
Production infrastructure has to be inspectable, portable, and close to the systems teams already operate. Standards and open tooling make AI-leveraged work safer.
Boring is a feature
Production software has to be fast, inspectable, reliable, measurable, and safe under real workloads. The future can be AI-leveraged without being theatrical.
Team
We are not new to this problem.
Tero was founded by Ben Johnson, creator of Vector.dev, and Nihar Singhal. Our team includes OpenTelemetry maintainers, creators of the telemetry policy standard, and engineers from Datadog, Lightstep, and New Relic. We have built the infrastructure, worked inside the platforms, and seen why the old model leaves teams with more evidence than understanding.
Open source
Creators of Vector.dev
Telemetry infrastructure adopted by production teams around the world.
Standards
OpenTelemetry maintainers
Maintainers in the ecosystem defining modern telemetry standards.
Policy
Creators of the telemetry policy standard
The standard behind portable telemetry control.
Experience
Former Datadog, Lightstep, and New Relic
Experience from the companies that shaped observability.
The future we are building toward
Production engineering for the AI era.
Evidence should become issues. Issues should move through the right action path. Fixes should be measured. Tero starts with logs.