Menu
Myriad: data extraction

Myriad extracts data from PDFs, scans, emails, and attachments. It turns messy documents into checked, structured data your teams and systems can use.

Built for teams that handle a lot of documents and need accurate results, clear checks, and a full record of what was extracted.

Core capabilities:
  • OCR and data extraction across different document layouts
  • One standard format for the extracted data
  • Checks against trusted reference data
  • Field-level traceability and confidence scores
  • API delivery or review screen
  • Human review and correction when needed
Myriad in use

The examples below show production use, proof-of-concept work, and internal demos.

The savings figures below use simple operating assumptions and include a 50% realism factor.

Pricing is indicative and is confirmed after a brief initial workshop.

Construction devis data extraction
What's the problem:
  • Construction companies receive supplier estimates in inconsistent formats - PDFs, Excel files, email bodies - requiring manual review, interpretation, and re-entry into internal systems.
  • This process is slow, error-prone, and difficult to standardize across projects and suppliers.

What Myriad does:
  • Myriad ingests incoming estimates, extracts key data (line items, quantities, pricing), and converts it into a structured format aligned with your internal systems.
  • It handles heterogeneous documents, normalizes supplier data, and flags inconsistencies automatically.

What's the impact:
  • Estimated efficiency savings of about CHF 30k per year (assumes 8,000 documents per year, 7.5 minutes saved per document, and CHF 60 per hour loaded cost)
  • Faster and more consistent estimates create a more professional experience for clients, suppliers, and internal reviewers, helping reduce friction before project approval.
  • More consistent validation and documentation
  • Faster turnaround for estimate checks and approvals
  • Foundation for automated pricing estimation

Typical engagement:
  • Initial set-up: 2-4 weeks depending on document variability
  • Indicative price for running MVP: Starting from CHF 10k per year (depends on volume and tailored rules)
  • Next step: send 1-2 sample documents. We'll assess feasibility and estimate potential impact.
Automate insurance application processing (PoC)
What's the problem:
  • Insurance providers handle a high volume of incoming application forms, often in mixed formats and layouts.
  • Manual data entry and validation are time-consuming, and seasonal peaks put additional strain on operations.

What Myriad does:
  • Myriad automatically parses application forms and attachments across heterogeneous formats.
  • It segments bundled PDFs into individual case documents, extracts key data, and delivers structured, validated case objects.
  • Output is delivered via API to underwriting systems or through a review interface when needed.

What's the impact:
  • Estimated efficiency savings of about CHF 50k per year for a mid-sized insurer (assumes 12,000 applications per year, 5 minutes saved per application, and CHF 50 per hour loaded cost)
  • Significant reduction in manual data entry and validation effort
  • Faster intake cycles and improved processing throughput
  • More consistent and reliable data for underwriting
  • Better initial experience for customers and brokers

Typical engagement:
  • Setup: ~1-3 months depending on document variability
  • Pricing: starting from CHF 0 (setup), ~CHF 15k/year (annual license)
  • Integration: API or simple batch processing (e.g. via SharePoint shared folders)
  • Next step: send 1-2 sample documents. We'll assess feasibility and estimate potential impact.
Claims document review automation (PoC)
What's the problem:
  • Claims teams manually review incoming documentation, often with inconsistent validation across reviewers.
  • Processing effort increases significantly during peak periods, slowing down follow-up and decision-making.

What Myriad does:
  • Myriad extracts key claim attributes from both structured and unstructured documents, validates completeness, and ensures required documentation is present.
  • It delivers structured outputs ready for downstream systems, with full traceability of extracted fields and associated confidence levels.

What's the impact:
  • Estimated efficiency savings of about CHF 40k per year (assumes 5,000 documents per year, 10 minutes saved per document, and CHF 50 per hour loaded cost)
  • Reduced manual review workload for targeted document types, with faster processing during peak periods
  • More consistent validation across claims, with improved claim data quality for downstream workflows
  • Reliable performance during peak volumes

Typical engagement:
  • Setup: ~1-3 months depending on document variability
  • Pricing: starting from CHF 0 (setup), ~CHF 10k/year (usage-based)
  • Integration: API or batch processing into existing systems
  • Next step: send 1-2 sample documents. We'll assess feasibility and estimate potential impact.
Document data extraction for logistics (demo)
What's the problem:
  • Logistics teams manually extract data from invoices and bills of lading, leading to incomplete processing and variable data quality.
  • This limits visibility and creates inefficiencies in downstream operations.

What Myriad does:
  • Myriad automatically extracts key shipment data - including carrier, receiver, delivery date, and package details - from incoming documents.
  • It provides structured outputs and reporting, ready for integration into ERP or tracking systems via API.

What's the impact:
  • Estimated efficiency savings of about CHF 20k per year (assumes 6,000 documents per year, 5 minutes saved per document, and CHF 50 per hour loaded cost)
  • Reduced manual data extraction effort, with more complete and reliable document processing
  • Improved data quality for operations and tracking, with better visibility across shipments
  • Consistent structured extraction across document layouts, creating a solid foundation for analytics and further automation

Typical engagement:
  • Setup: ~1-3 weeks depending on document variability
  • Pricing: starting from CHF 0 (setup), ~CHF 5k/year (usage-based)
  • Integration: API or batch processing into ERP / tracking systems
  • Next step: send 1-2 sample documents. We'll assess feasibility and estimate potential impact.
Data validation and enrichment
What's the problem:
  • Operational processes often rely on incomplete or inconsistent reference data such as names, addresses, and legal entities.
  • Manual validation slows down processing, introduces errors, and leads to inconsistent decisions across teams.
  • Limited traceability also makes it difficult to understand why a value was accepted or rejected.

What Myriad does:
  • Myriad validates extracted data against trusted sources and enriches it with missing or additional attributes.
  • It checks fields, adds relevant information such as entity identifiers or country risk, and keeps full traceability of what was validated, enriched, and where the data originates from.
  • Results are delivered via API or through a review interface.

What's the impact:
  • Estimated efficiency savings of about CHF 15k per year (assumes 20,000 checks per year, 1.5 minutes saved per check, and CHF 60 per hour loaded cost)
  • Reduced manual validation effort, with more consistent and reliable decision-making
  • Improved data quality across systems, with full traceability of validation and enrichment steps
  • Improved consistency of reference data across workflows, reducing rework and strengthening downstream automation and decision systems

Typical engagement:
  • Setup: immediate for pilot, depending on data sources
  • Custom connectors: ~1-2 months
  • API and secure integration: ~1-2 months
  • Deeper system integration: depends on target systems
  • Next step: send 1-2 sample datasets or use cases. We'll assess feasibility and estimate potential impact.
Powered by sense6