How It Works
MidLyr has two core API surfaces:
- Regulation API — access curated regulatory source material.
- Analysis API — evaluate submitted content using regulatory and company-specific context.
Regulation API
Section titled “Regulation API”The Regulation API gives your agents and applications access to a curated compliance-source library.
1. We ingest public regulatory sources
Section titled “1. We ingest public regulatory sources”MidLyr collects public compliance sources from many upstream locations, including statutes, regulations, interagency and agency guidance, examination handbooks, interpretive actions, SRO rules, and related materials.
These sources can come from different agencies, jurisdictions, formats, and publishing workflows. MidLyr handles the ingestion and normalization work so your team does not have to maintain a separate retrieval pipeline.
2. We normalize and classify the corpus
Section titled “2. We normalize and classify the corpus”After ingesting source material, MidLyr normalizes and classifies it into a consistent API model.
This includes:
- assigning stable document IDs
- identifying the source category
- tracking authorities and jurisdictions
- generating useful descriptions
- organizing content for browse, read, and query workflows
The result is a consistent set of Regulation Objects and Regulation Detail Objects across source types.
3. We build retrieval-ready context
Section titled “3. We build retrieval-ready context”MidLyr prepares the regulatory corpus so callers can:
- browse document-level matches
- read a single document’s metadata and table of contents
- fetch document text by offset, TOC section, or cursor
This powers the Regulations API endpoints:
Analysis API
Section titled “Analysis API”The Analysis API evaluates submitted content using both regulatory context and your company-specific context.
1. We build your company context
Section titled “1. We build your company context”When your organization signs up, MidLyr constructs a context profile for your company. This context may include information such as:
- your business model and product surface
- operating jurisdictions
- relevant regulatory categories
- internal policies or procedures you provide
- review standards or escalation preferences
This context helps MidLyr avoid treating every analysis request as a generic regulatory question.
2. We retrieve relevant regulatory context
Section titled “2. We retrieve relevant regulatory context”For each analysis request, MidLyr can use the regulatory library to find relevant source material from the curated corpus.
Findings point back to source text with document-level provenance.
3. We combine company context with request context
Section titled “3. We combine company context with request context”When you submit an analysis request, MidLyr evaluates the submitted content against both:
- the dynamic company context for your organization
- the relevant regulatory source context for the request
For example, a sentence that is acceptable for one product, jurisdiction, or customer scenario may be risky in another. The company context helps the analysis reflect that difference.
4. We return structured findings
Section titled “4. We return structured findings”The Analysis API returns structured output rather than a free-form answer. For a screening job, the result includes:
riskScore— overall compliance risk score (0–100)findings— list of compliance findings, each with a priority level (p1/p2/p3), details, and regulatory text references with document-level provenance
See Screen Analysis API for the full response shape.
Customize Your Context
Section titled “Customize Your Context”If you want to define, review, or edit your company context, contact MidLyr. We can help configure the context profile used for your analysis requests.
This is especially useful if your institution has custom policies, product-specific rules, jurisdiction-specific review standards, or internal escalation requirements.