Ukraine Rail Freight Logistics Email & Document Archive — Anonymised Business Correspondence, SMGS Consignment Notes, Invoices | 2014–2022 | RU/UK/EN
Overview
Ukraine Rail Freight Logistics Email & Document Archive — Anonymised Business Correspondence, SMGS Consignment Notes, Invoices | 2014–2022 | RU/UK/EN is an audio and text dataset. Typical applications include document understanding & OCR. Relevant sectors: AI & Machine Learning, Automotive & Mobility and Manufacturing & Industrial. It is offered as a paid dataset; pricing details are provided on request.
Primary applications
The complete operational email archive of a Ukrainian rail-freight logistics company — approximately 1.2 TB of correspondence and the documents attached to it, including files recovered from accounting software, covering the business from 2014 until 2022, after which the operations gradually declined. The company both rented out ints own wagons, and well as operated railway’s ones. It also repaired wagons and worked under annual contracts with DTEK from 2014 and served as SOCAR's exclusive rail freight operator. Email threads with their reply chains, and the documents attached: SMGS consignment notes, tariff schedules, reports of completed services, invoices, wagon manifests&tracking, schedules, etc. The threads and the attachments are linked, so a shipment can be followed from the first enquiry to the final settlement. A lot of the documents are scanned rather than digital — photographed notes with stamps, signatures, handwritten corrections and natural degradation of office equipment. Those pages have been recovered via OCR. The dataset also includes 1C and M.E.Doc archives, which are softwares that were used for document filling and full accounting of the company. Wagon numbers, consignment and invoice numbers, station and route names, tonnages, tariffs, monetary amounts and contract dates are all retained. That operational detail is the strong suit of this dataset and stayed preserved during de-identification. This kind of business operational correspondence is largely ab
Technical specifications
- Modalities
- Audio, Text
Related datasets
More in Speech & Audio →Multimodal Data Engine: Annotation & Curation for High-Value AI (Polish)
The provider delivers a unified data development platform for creating production-ready datasets across the most demanding AI domains, including Physical AI (robotics,…
- Language
- Polish
- Volume
- —
- Domain
- Technology & IT +5
- Use case
- Fine-tuning & instruction data
More speech & audio datasets
All 1,727 speech & audio datasets →- Ukrainian conversational audio, channel-separated speakers
- Ukrainian conversational audio, live single-channel
- Ukrainian conversational podcasts across genres
- Ukrainian conversational podcasts, various genres
- Ukrainian live conversational audio, channel-separated
- Ukrainian Speech-to-Text Transcribed Audio Dataset
- United States City Environmental Data | Street Noise Levels | CCPA Compliant | 100% Traceable Consent
- Unscripted Call Center Telephony Speech Data | 20,000 Hours | Speech Recognition Data | Speech AI Datasets | Audio Data
- Unsupervised Spontaneous French Speech Data | 10,019 Hours | Large Language Model (LLM) Data