teiō
“What does each person here do every week that a skill could draft?”

A skills platform, so every team automates its own work.

Most weekly work repeats. A skill writes one job down so an AI produces the first draft inside the tools people already use, and the platform vets, grades and counts those skills.

What we build

Your peopleWhat we build
RolloutWorkshops on your examplesWeekly office hoursUsage by teamRecognition for authors

Three AI judges score each version against its rubric. A dimension passes at a mean of 7 with no judge below 5.

Palette and logo8.5
Cover and contents7.7
Layout8.2
Writing style7.3
passed · v3 · approved

Fifty skills, built by engineers and HR alike.

Everyone from engineers to people ops has shipped one. Release notes, feature audits and bug audits are themselves skills.

The teiō skills store.

Skills
What your Claude can already do. Say the words, or type the command.
512 uses in 30 days · 52 skills · 14 creators
Search skills, or describe a task
Sales
Create proposal
Build a proposal from a call and the research, in the house format
Tarush · 41 uses
Proposal follow-up
Diagnose a quiet proposal and draft the next move
Tarush · 22 uses
Data engineering
PII candidate identification
Find and confirm personal-data columns in any warehouse
Sagar · 9 uses
Snowflake to BigQuery
Convert SQL between the two, either direction
Shridhar · 6 uses
People
Onboarding plan
Provision tools before a start date, one plan to approve
Priya · 11 uses
Weekly goals
Turn last week's record into this week's three goals
Chitra · 58 uses
Content
Brand a deliverable
Word, PDF, deck or sheet in the house style
Prince · 34 uses
Weekly story
Draft the weekly story with the six-field fact strip
Tarush · 7 uses
Brand a deliverable · v3
Submitted by Priya (HR) · improvement on v2
SCORING · 3 judges
DimensionAnthropicOpenAIGoogleMean
Palette and logo0.00.00.00.0
Cover and contents0.00.00.00.0
Layout0.00.00.00.0
Writing style0.00.00.00.0
Pass rule: a dimension passes at a mean of 7 with no judge below 5. No score may drop more than a point against v2. Safety and duplicate review ran first.

Every version is scored before it ships.

A skill improvement from people ops, scored by three judges. Pass at a mean of 7, no judge below 5, no dimension down more than a point against the version before.

Keep your skills in a repository you control, not one vendor's store.

Claude, ChatGPT, Gemini and Copilot all have skill libraries. None reaches across the others, usage tracking is patchy, and when the best model changes you rebuild.

Ownership

Every skill in one git repo you control. You decide which are on and who gets them.

One count

A usage hook fires on every run in every tool and reports to your platform. Adoption by department, finally measurable.

Quality that rises

Weak first attempts (2s and 4s) reach 7 to 9 within a few revisions because authors see every score and its reasons.

Day-one skills

New joiners are pointed at the skills for their role on day one. Onboarding becomes a list, not a month.

Case studies

/ Property and casualty insuranceSoutheast US

A P&C insurer went from stored procedures to a warehouse, then to AI across the company

Data platform in production
Before
Logic buried in SSIS jobs, one data engineer
After
Fivetran, Snowflake and dbt in production, four hires, Claude rolled out
Data intelligence

The other three practices

Agentic engineeringEngineers embedded with your team, building the software your company runs onData intelligenceTrusted answers from your own data, then agents that reason over itAI strategyA forward-deployed AI strategist, a day a week, who finds the AI projects and defines them

Own the AI that runs your company.

Built inside your company and yours to keep. Thirty minutes with a practice lead.

Book a first call