The Chosen Reference Standard
AI should follow the same rules your business does.
Lectotype applies your policies, permissions, regulatory requirements and approval rules before AI acts. Whichever model you use, those rules still apply.
When AI Takes Action
AI can do more than answer a prompt.
It can recommend a treatment, approve a request, adjust a threshold or influence what happens next. Before that work moves forward, an organization may need to know what the AI can access, which rules apply and when human approval is required.
Lectotype brings those checks into the request path before the model responds.
How Lectotype Works
The rules are checked before the request reaches the model.
Lectotype sits between the request and the model. It checks access and permissions, applies the policies and regulatory requirements that govern the work, and determines whether approval is needed. The request can then move forward or be routed for review.
01
Request
User, system, or agent
02
Lectotype
Rules checked before the request continues
- Access and permissions
- Applicable policies and regulatory requirements
- Approval paths and human review
03
Model
OpenAI · Anthropic · Gemini · Proprietary models
04
Response or Action
Moves forward after the applicable checks are complete
Rules first. Intelligence second.
OpenAI, Anthropic, Gemini and in-house models can all sit behind the same Lectotype control layer, so changing providers does not require rebuilding governance around each one.
Supported Models & Integrations
Standard
Eight pillars for governing AI.
STANDARD is Lectotype’s framework for evaluating governance across AI use. Its eight pillars cover data sovereignty, traceability, assurance, built-in governance, deployment safety, authority, regional compliance and dependability.
Why Lectotype?
In taxonomy, a lectotype is the specimen chosen as the definitive reference for a species. The name reflects the role Lectotype is designed to play for enterprise AI: a consistent reference for the rules that apply as models and requirements change.
Eight pillars
Where It Applies
Built for decisions that carry real consequences.
Approving a transaction, recommending a treatment or adjusting a safety threshold involves more than producing a useful answer. The decision may also need to satisfy internal policy, regulatory requirements, jurisdictional rules and human approval.
A global pricing decision may involve several jurisdictions at once. Lectotype checks the relevant requirements before the decision moves forward.
Results in Practice
Results from regulated environments.
National RIA
Financial Services
$1.7T
in assets under management
3×
advisor capacity
60%
reduction in compliance documentation
—
Signed on to expand
Prime42
Water Intelligence
Live
in production
5×
report output
Token Cost Calculator
What is your AI usage costing?
Estimate current token spend and compare how different model-routing assumptions affect cost.
Editable. Use your own contracted rate.
Routed work is priced at the mid-tier rate.
Estimated monthly spend
$15,000
Estimated monthly difference from routing
$4,800
Figures are arithmetic on the values above, shown for estimation only. Rates and routing share are yours to set.
See how Lectotype would handle your use case.
Bring us the model, policies and approvals involved. We’ll walk through the request path with you.
Corporate and IP counsel: Wilson Sonsini Goodrich & Rosati