Eng Health Check — Jeremy Rajan

Score your engineering organisation across 6 key dimensions. Takes about 3 minutes. No sign-up required.

1 of 6

Delivery

How often do you deploy to production?

Monthly or less

Every 1-2 weeks

A few times per week

Daily

Multiple times per day, on demand

What's your lead time from commit to production?

More than a week

A few days

About a day

A few hours

Under an hour

Do you have CI/CD pipelines?

No, deployments are manual

CI only (automated tests), manual deploys

CI/CD for some services

CI/CD for most services

Fully automated CI/CD across all services

2 of 6

Quality

How do you approach testing?

Little to no automated testing

Some unit tests, mostly manual QA

Good unit test coverage, some integration tests

Comprehensive unit + integration tests, CI enforced

Full test pyramid with e2e, contract tests, and quality gates

What's your code review process?

No formal review process

Optional reviews, sometimes skipped

Required reviews, but often rubber-stamped

Thorough reviews with clear standards

Strong review culture with automated checks and guidelines

Can you roll back a bad deployment quickly?

No rollback strategy, we fix forward manually

Rollback is possible but slow and manual

Rollback within 30 minutes for most services

Quick rollback with feature flags or blue-green deploys

Instant rollback, canary releases, automated rollback on errors

3 of 6

Observability

What's your logging setup?

Console logs, SSH to check them

Basic log aggregation but unstructured

Centralised structured logging for most services

Structured logging with search and correlation IDs

Full observability: logs, traces, and metrics correlated

How do you find out about production issues?

Users report them

Someone checks dashboards manually

Basic alerts, but noisy / often ignored

Targeted, actionable alerts with clear ownership

Proactive detection with anomaly alerts and SLO-based monitoring

Do you have dashboards that answer business questions?

No dashboards

Default cloud provider dashboards only

Custom dashboards for some services

Comprehensive dashboards covering key business metrics

Real-time dashboards with drill-down, shared across eng and product

4 of 6

Team Health

Is there clear ownership for every service and domain?

No, it's unclear who owns what

Informal ownership, depends on who built it

Most services have a team assigned, some gaps

Clear ownership with documented responsibilities

Full service catalogue with owners, SLAs, and escalation paths

Do you have an incident management process?

No process, whoever notices it fixes it

Ad-hoc response, no postmortems

Basic on-call rotation, some postmortems

Defined incident process with severity levels and blameless postmortems

Mature incident management with runbooks, war rooms, and trend tracking

How is your engineering documentation?

Almost no documentation

Some docs but mostly outdated

Key systems documented, kept reasonably current

Comprehensive docs with ADRs and onboarding guides

Living documentation culture — docs are part of the definition of done

5 of 6

Architecture

How coupled are your services?

Monolith or tightly coupled services, shared database

Some services separated but many shared dependencies

Mostly independent services, a few shared resources

Independent services with clear API contracts

Fully decoupled with event-driven communication where appropriate

Can teams deploy independently without coordinating with other teams?

No, most changes require coordination across teams

Sometimes, but cross-team coordination is frequent

Most teams can deploy independently for most changes

Teams are autonomous, coordination is rare

Fully autonomous squads with self-service platform

How do you handle scaling and performance?

React to outages, scale manually when things break

Some auto-scaling, but no capacity planning

Auto-scaling for key services, basic load testing

Proactive capacity planning with regular load testing

Elastic scaling with performance budgets and chaos engineering

6 of 6

AI Readiness

How well-documented is your codebase and architecture?

Tribal knowledge only, very little written down

Some READMEs and scattered docs

Key systems documented, architecture diagrams exist

Comprehensive docs, ADRs, and up-to-date Confluence/Notion

Documentation is searchable, structured, and part of the workflow

Are engineers using AI-assisted development tools?

No AI tooling in use

A few individuals experimenting on their own

AI coding assistants available but no guidelines

AI tools adopted with team guidelines and best practices

AI integrated into workflows with centralised skill definitions and guardrails

Do you have a centralised knowledge base that AI tools could query?

No, knowledge is scattered across Slack, email, and people's heads

Some Confluence/Notion pages but poorly organised

Organised knowledge base but not searchable by AI

Structured knowledge base with good search

AI-ready knowledge layer: indexed, searchable, connected to dev tools

Engineering Health Check

jeremyrajan.com/eng-health-check · Generated

out of 30

Recommendations

Want to talk through your results and identify the highest-impact improvements?

Book a free 30-min session →