Skip to content
novelavia

Product

From freight documents to actionable intelligence.

Novelavia is an AI-powered document intelligence and workflow automation platform built specifically for freight and transportation operations. It reads the documents that move freight and turns them into structured, validated data.

The load lifecycle

Designed to support every stage of a load

The long-term vision is an intelligent automation layer that connects documents, dispatchers, drivers, brokers, accounting systems and transportation management systems. The MVP starts where every load starts: the rate confirmation.

  1. 01Load bookingLoad details agreed with the broker
  2. 02Rate ConfirmationMVP focus: extraction and validation
  3. 03Dispatch preparationDriver instructions
  4. 04PickupBOL, check-in, seal numbers
  5. 05TransportationTracking and exceptions
  6. 06DeliveryAppointments, lumper, detention
  7. 07Proof of DeliverySignatures and exceptions
  8. 08InvoicingInvoice packet and accessorials
  9. 09Payment reconciliationTerms, quick pay, payments

Highlighted: current MVP focus. Every other stage is a planned expansion of the platform.

Core features

Five capabilities, one platform

Each capability is labeled with where it stands today, so it is always clear what you can try now and what is still on the roadmap.

01Proof of concept · MVP focus

AI Rate Confirmation Processing

Designed to read rate confirmations and return structured shipment data, with a confidence score for every field.

  • Load number, broker and carrier details, MC and USDOT numbers
  • Agreed rate, currency, line items and payment terms
  • Pickup and delivery locations, dates, times and reference numbers
  • Commodity, weight, equipment type and number of stops
  • Detention terms, extra charges and special instructions

Today

Interactive demo with labeled sample results. A Gemini-based extraction pipeline is implemented and can be switched on per deployment.

Next

Accuracy benchmarks on anonymized documents, broker-specific layouts and batch processing.

02In development

Intelligent Document Classification

Recognizes what each document is before extracting anything, so rate confirmations, BOLs, PODs and invoices each get the right treatment.

  • Rate Confirmation, Bill of Lading and Proof of Delivery
  • Freight Invoice and Carrier Agreement
  • Lumper Receipt and Fuel Receipt
  • Unrecognized documents are flagged instead of forced into a type

Today

Classification is part of the extraction pipeline and shown in the demo results.

Next

Mixed-document batches, multi-document PDFs and per-type accuracy reporting.

03Proof of concept

AI Document Validation

Business rules check every result before anyone relies on it. Uncertain or inconsistent values are routed to a person, and missing values are reported instead of guessed.

  • Missing information and incomplete addresses
  • Conflicting dates and rate totals that do not add up
  • Missing signatures and missing delivery documentation
  • Discrepancies between rate confirmations and invoices
  • Low-confidence fields highlighted for human review

Today

Deterministic validation rules run in the demo, including a cross-check between a sample invoice and its rate confirmation.

Next

Duplicate detection, customer-specific rules and validation across a load's full document history.

04Planned

Automated Dispatch Preparation

Turns verified load data into clear driver instructions: stops, appointment times, equipment, reference numbers and special requirements.

  • Pickup and delivery details with appointment windows
  • Equipment requirements and reference numbers
  • Special instructions, PPE and tracking requirements
  • Broker contact details for check calls

Today

The demo includes a template-based preview of a driver instruction sheet built from reviewed fields.

Next

Delivery to drivers through messaging and dispatch tools, with confirmation tracking.

05JSON and CSV export available · integrations planned

Document-to-Workflow Automation

Verified data is exported as JSON or CSV today. Direct delivery to the systems your team already uses is planned, so nobody has to type a load in twice.

  • JSON and CSV exports (available in the demo)
  • Transportation management systems (TMS)
  • Accounting systems and dispatching platforms
  • Fleet management tools and internal APIs

Today

Reviewed results can be downloaded as structured JSON or CSV.

Next

Native TMS and accounting integrations, webhooks and a documented REST API.

Trustworthy automation

AI that knows when to ask a person

Freight data drives payments, appointments and compliance. Novelavia is built to be transparent about uncertainty instead of confidently wrong.

  • Reports gaps instead of guessing

    Extraction is instructed to return only values printed on the document. Missing required fields are flagged for review, not filled in.

  • Confidence on every field

    Each extracted value carries a confidence score, so reviewers know exactly where to look.

  • People review what matters

    Uncertain, missing and inconsistent values are routed to a person. Every edit is tracked against the original.

  • Documents are data, not instructions

    Document text is treated as data. Safeguards against prompt injection keep it from changing how Novelavia behaves, and the model has no actions to trigger.

Exports and integrations

Verified data, wherever you need it

Structured exports are available in the demo today. Direct integrations with the systems freight teams use every day are part of the roadmap.

  • JSON exportAvailable
  • CSV exportAvailable
  • Transportation management systems (TMS)Planned
  • Accounting systemsPlanned
  • Dispatching platformsPlanned
  • Fleet management toolsPlanned
  • Internal APIs and webhooksPlanned
  • Email ingestionPlanned

Roadmap

What exists today and what comes next

Novelavia is in early development. The proof of concept is live in the demo; the MVP is being built now.

  1. Phase 1Available in the demo

    Proof of Concept

    • Rate Confirmation processing
    • Structured data extraction
    • Document preview
    • Validation rules
    • JSON and CSV export
  2. Phase 2In development

    MVP

    • Real AI-powered document analysis
    • Multiple document types
    • User authentication
    • Document history
    • Batch processing
    • Human review workflows
  3. Phase 3Planned

    Intelligent Automation

    • Email ingestion
    • Automated load creation
    • Driver dispatch instructions
    • Cross-document validation
    • TMS integrations
    • Accounting integrations
  4. Phase 4Planned

    AI Logistics Intelligence

    • Operational analytics
    • AI-assisted dispatch decisions
    • Proactive exception detection
    • Automated document reconciliation
    • Intelligent workflow orchestration

Help shape Novelavia from day one.

We are looking for U.S. carriers, brokers, dispatch services and freight accounting teams to pilot the MVP. Early access members try new capabilities first and directly influence the roadmap.