DokuBasis: Create Structured Reports Faster with Voice and AI

People who observe, inspect, discuss, or document things while on the move rarely have both hands free for extensive text entry. As a result, notes often remain incomplete, are not transcribed until hours later, or have to be pieced together from several voice memos, photos, and bullet points.

DokuBasis turns this into a seamless workflow: Record speech directly in the browser or upload existing audio files, transcribe them automatically, and use AI to assign the identified information to a suitable report template. The result can be reviewed, supplemented, documented with photos, and then exported as a PDF or Word file.

DokuBasis is therefore more than an audio-to-text service. The web app transforms spoken information into an editable, structured AI report - for example, a digital construction log, a care report, meeting minutes, or a property survey.

What is DokuBasis?

DokuBasis is a browser-based application for AI-assisted report creation. It is designed for people who need to capture information directly on site or during a conversation and turn it into clear, traceable documentation.

Instead of typing every item into a form individually, users dictate their observations, actions, tasks, or results. DokuBasis then performs two steps:

1. The recording is converted to text using Azure Whisper.
2. Azure OpenAI analyzes the text and assigns the information to the fields in the selected template.

For example, the sentence “The company Beispielbau was on site today with six employees and began drywall work on the second floor” can be used to generate information about the company, workforce size, location, and work performed. Which information is identified and where it is assigned depends on the selected report template.

The automatically generated content remains editable. Users can correct details, add missing information, and review the report for technical accuracy before exporting it.

Who can benefit from DokuBasis?

DokuBasis is particularly suitable for activities where a large amount of information is generated in a short period of time and structured documentation is required.

Construction managers, site supervisors, and construction inspectors

On a construction site, weather, companies present, workforce sizes, work performed, obstructions, instructions, and acceptances must be documented in a traceable manner. With DokuBasis, this information can be dictated directly on site and organized into a construction log in accordance with VOB. Photos can additionally document conditions, construction progress, and defects.

Home care professionals and care services

Care professionals can record observations, vital signs, activities performed, medication, pain, and any need for medical consultation. Voice input reduces typing while on the move. The generated home care report must be reviewed and supplemented if necessary before further use.

Real estate agents, appraisers, and property inspectors

During an inspection, property features, rooms, fixtures, conditions, and defects can be dictated directly. DokuBasis structures the information as a property survey, helps create a room schedule, and collects content that can later be reused for a property description or listing brochure.

Teams, project managers, and minute takers

Meetings produce decisions, tasks, responsibilities, and deadlines. DokuBasis helps consolidate this content into meeting minutes instead of having to reconstruct it later from disorganized notes.

Safety professionals and security services

Safety inspections, patrols, incidents, identified defects, and resulting actions must be documented in a traceable manner. For this purpose, DokuBasis provides dedicated templates for safety professionals as well as for building security and security services.

Facility managers and building management

Maintenance work, technical issues, damage, inspections, and outstanding tasks can be captured by voice and organized in a facility management report. Photos supplement the documentation directly within the relevant record.

Self-employed professionals and general office work

Not every use case requires an industry-specific structure. A general report and a Daily Journal are available for daily reports, free-form documentation, or personal work journals.

Why is an AI report more than a transcription?

A conventional transcription renders spoken content as continuous text. The actual work often only begins afterward: names, dates, tasks, responsibilities, and technical details must be identified and manually transferred to a form.

DokuBasis, by contrast, combines transcription and structured information extraction. The selected template specifies which fields are required. The AI attempts to assign relevant information from the spoken content to these fields.

This offers several practical advantages:

• less manual typing on site,
• a consistent structure for recurring documentation,
• easier capture of tasks, participants, deadlines, and defects,
• direct editing in a structured form,
• combined output of text and photos,
• ready-to-use export as PDF or DOCX.

AI supports the data capture process, but it does not replace professional judgment or final review. Particularly for safety-related, medical, contractual, or legally significant content, the result must be reviewed by a qualified person.

How does DokuBasis work?

1. Select a report template

First, select the appropriate template. Each template contains its own form fields and instructions for the AI. A construction log requires different information than a property survey or care report.

2. Record speech or upload audio

Speech can be recorded directly using the microphone on a smartphone, tablet, or computer. Alternatively, existing files in MP3 or M4A format can be imported. Multiple recordings can be combined for processing.

3. Transcribe the recording

The audio data is technically processed and temporarily transferred to Azure Whisper within the EU. Whisper converts the recording into written text.

4. Structure content automatically

Azure OpenAI evaluates the transcript based on the selected template. Identified information is transferred to the appropriate form fields. This may include participants, activities, deadlines, defects, actions, responsibilities, or property features.

5. Review and supplement the result

The AI-generated result can be edited directly in DokuBasis. Users should check spelling, numbers, technical context, and completeness. Information not yet mentioned can be added manually.

6. Add photos

Photos can be taken directly with the device camera or selected from existing images. This makes it possible to visually document defects, construction progress, technical conditions, or property features, for example.

7. Export as PDF or Word

The reviewed report, including photos, can be downloaded as a PDF or DOCX file for Microsoft Word. Empty fields are omitted during export to create a clear document.

Which reports does DokuBasis support?

DokuBasis currently includes nine templates for different industries and documentation tasks.

Digital construction log in accordance with VOB

The construction log template supports structured construction site documentation. Instead of transferring companies, workforce sizes, trades, and special events from notes later, this information can be dictated directly on the construction site. The template supports documentation but does not automatically guarantee the legal or contractual completeness of a construction log. The requirements of the respective project and contract remain decisive.

Home care report

The care template assigns spoken observations and actions to the appropriate fields. This allows information to be captured directly within the workflow. DokuBasis is a documentation tool, not a medical device, diagnostic system, or substitute for required specialist software and professional decisions.

Property survey and room schedule

The real estate template helps collect information room by room during a property inspection. In addition to conditions and defects, features for a later property description can also be recorded. Images can be exported together with the structured information.

Meeting minutes

The meeting template focuses particularly on outcomes: What was discussed? What decision was made? Who will take on which task, and by when? This turns spoken notes into minutes that are easier to share and track.

Safety, security service, and facility management reports

These templates are suitable for recurring inspections, walkthroughs, and handovers. Observations can be recorded directly at the relevant location. Together with photos, they provide a traceable basis for internal coordination and further action.

Key features at a glance

Direct voice recording in the browser

DokuBasis runs as a web app. No conventional desktop installation is required for recording. Once microphone permission has been granted, users can dictate directly in the browser. The recording can be stopped manually, ends automatically after an extended period of silence, or stops after ten minutes at the latest.

Multiple audio recordings and file import

A report can consist of multiple voice recordings. Existing MP3 and M4A files can also be used. This is helpful when content has been captured at different locations or at different times.

AI glossary for technical terms and proper names

Company names, personal names, product names, and technical terms are not always spelled correctly by automatic transcription. Preferred spellings can be stored in the DokuBasis glossary. These notes are incorporated into transcription and structuring and can improve the recognition of project-specific terms.

Structured, editable forms

Depending on the template, suitable input fields are available, including text, numbers, date, time, selection fields, toggles, and multiple selections. Automatically identified values can be changed afterward.

Photo documentation

Photos can be taken with the camera or selected from the gallery. DokuBasis reduces the size of large images for further processing and can include them in the PDF or Word export if desired.

Local storage in the browser

Reports, form values, recordings, and added media are stored persistently in the local browser database on the device being used. DokuBasis cannot access this locally stored content.

At the same time, this local storage means that the data is tied to the browser and device being used. If browser data is deleted, a different browser is used, or another device is used, locally stored reports will not automatically be available there. Important documents should therefore be exported in good time and backed up in accordance with your own requirements.

PDF and Word export

After review, a report can be downloaded as a PDF or DOCX file. PDF is particularly suitable for a fixed format that is easy to share. DOCX enables further editing in Microsoft Word or compatible word processing programs.

Which formats and technical methods are supported?

At DokuBasis, “supported reports” usually refers to the document templates described above. The following formats and methods also apply to technical data exchange:

• Voice input: direct microphone recording in the browser,
• Audio import: MP3 and M4A,
• Images: camera capture or selection of local image files,
• Document export: PDF and DOCX,
• Sign-in: Google or Microsoft via OAuth 2.0/OpenID Connect,
• Data transmission: encrypted via HTTPS/TLS,
• Local data storage: IndexedDB in the browser.

The following limits currently apply to imported audio files:

• a maximum of 20 audio files per processing operation,
• a maximum of 20 MB per file,
• a maximum of ten minutes per audio file,
• WAV files are not supported.

An internet connection is required for transcription, AI structuring, sign-in, and document export. Local storage does not make DokuBasis a fully offline application.

Data privacy and data sovereignty

DokuBasis follows an approach in which the reports themselves are stored persistently on the end device.

What remains on the device?

Reports, form values, audio recordings, notes, and images are stored persistently in the browser’s IndexedDB. This means that this content is not stored as a central collection of reports on DokuBasis servers.

What is processed temporarily?

For transcription and AI-assisted structuring, audio and the content derived from it must be processed temporarily. For this purpose, DokuBasis uses Azure OpenAI services in the Sweden region within the EU. According to the information in the privacy policy, this data is not stored there permanently and is not used to train the AI models.

When a PDF or Word document is created, the required report content is briefly transferred to the DokuBasis backend. The backend is hosted by Hetzner in Germany. The content is processed to generate the export and is not subsequently stored permanently on the server.

All transmissions are encrypted via SSL/TLS. Detailed and legally authoritative information is available in the privacy policy. The data processing agreement is also available for review.

What DokuBasis does not handle automatically

DokuBasis is designed to simplify documentation work, but it does not promise error-free or legally compliant automation.

• AI output may be incomplete or incorrect.
• Names, numbers, deadlines, and technical statements must be checked.
• The app does not replace medical, legal, safety-related, or professional assessment.
• Users remain responsible for the content, completeness, retention, and lawful processing of their documents.
• Industry-specific requirements, company policies, and statutory retention obligations must still be observed.

This clear division of responsibilities is precisely what makes its use effective: DokuBasis handles time-consuming preparatory work, while professional review remains in human hands.

Try DokuBasis for free

You can get started directly in the browser. Sign-in is available with a Google or Microsoft account. Five free reports are available to try.

1. Open DokuBasis.
2. Sign in with Google or Microsoft.
3. Select a suitable template.
4. Record speech or upload an MP3/M4A file.
5. Review, supplement, and export the AI report.

Less typing afterward, more structure in your documentation: Create five free reports with DokuBasis now.

Frequently asked questions about DokuBasis

What exactly does DokuBasis do?

DokuBasis transcribes speech or audio files and automatically assigns identified information to the fields of a selected report template. The result can be edited, supplemented with photos, and exported as PDF or DOCX.

Is DokuBasis only transcription software?

No. A basic transcription produces continuous text. DokuBasis also uses AI to assign information such as activities, participants, tasks, deadlines, defects, or actions to structured form fields.

Which reports can I create with DokuBasis?

Templates are currently available for a construction log in accordance with VOB, a home care report, a property survey, meeting minutes, a safety inspection, building security and security services, facility management activities, a Daily Journal, and a general report.

Can DokuBasis create a digital construction log?

Yes. The construction log template captures weather, companies, workforce sizes, trades, work performed, obstructions, instructions, and acceptances, among other information. The result must be reviewed and supplemented in accordance with the requirements of the respective construction project.

Is DokuBasis suitable for care documentation?

DokuBasis offers a template for home care reports. It supports voice-based capture and structuring, but it does not replace professional review, required care documentation systems, or medical decisions.

Can I transcribe existing audio files?

Yes. In addition to direct microphone recording, the interface supports MP3 and M4A files. Multiple recordings can be added to one report.

Are WAV files supported?

No. WAV files are not currently supported. Use MP3 or M4A for file imports.

How long can a recording be?

An individual recording or imported audio file can be no more than ten minutes long. Up to 20 audio files can be used per processing operation. The maximum file size is 20 MB per file.

Does DokuBasis recognize company names and technical terms?

An AI glossary is available for names, companies, and technical terms. Spellings stored there provide additional context for transcription and structuring. Error-free recognition still cannot be guaranteed.

Can I include photos in a report?

Yes. Photos can be taken directly with the camera or selected from existing images. They can be exported together with the report as a PDF or Word file.

Can I edit the AI-generated result afterward?

Yes. All automatically populated form fields can be reviewed, corrected, and supplemented. This review is explicitly part of the intended workflow.

Which export formats are available?

DokuBasis creates PDF documents and DOCX files. Photos can be included in both export formats.

Where does DokuBasis store my reports?

The web app stores reports, form values, audio, and images persistently and locally in the IndexedDB of the browser being used. Data required for AI processing and export is transferred temporarily. The privacy policy provides further details.

Can I synchronize my reports across multiple devices?

Automatic synchronization of locally stored reports between devices is not currently part of the described feature set. Local data is therefore not automatically available on another device or in another browser.

Does DokuBasis work offline?

Reports are stored locally in the browser. However, sign-in, transcription, AI structuring, and export require a connection to the DokuBasis backend or the services used. A fully offline workflow is therefore not available.

Is DokuBasis free?

After signing in with Google or Microsoft, five reports can be tried free of charge. Information about additional plans is displayed within the app.

Does an AI report need to be reviewed?

Yes. Automatically generated content may contain errors or omissions. The report must be reviewed by a suitably qualified person before use and corrected if necessary.

Conclusion: Capture speech, structure content, share the report

DokuBasis bridges the gap between a quick voice memo and usable documentation. The app combines voice input, AI transcription, automatic structuring, industry-specific templates, photos, and document export in a single workflow.

This is particularly helpful wherever information is generated directly on site: on construction sites, during care visits, in real estate, at meetings, on security patrols, or in building management. Users save some of the effort of transferring information afterward while retaining control over the finished result.

Spoken observations become a structured, editable report.