Construction sites generate a constant stream of decisions, instructions, photographs, reports and unresolved defects. Gewerkton starts with a direct response to that reality: let people record the work by speaking while they are still on site, then turn that dictation into structured evidence.
Gewerkton · platform anatomy
From scaffold
to structured evidence
A German-rooted, globally framed construction platform turns spoken records at the point of work into evidence that can travel across plans, models, systems and regions.
“On site, what counts is what’s proven.”
The compact operation behind the global scope
languages
Multilingual by design
EU, US and APAC participants can work in their own languages from capture through report.
providers
Bring your own AI
Users bring their own keys and select providers across the EU, US and Asia, including mainland China.
in one night
An agent-assisted build
One solo founder directed Codex and Claude; delivery was checked with negative controls and mutation tests.
clips & posters
A broad public surface
The marketing site spans 27 languages, uses zero trackers and has no cookie banner.
One branded house · three connected product lines
Voice-first capture for evidence, defects, daywork reports, takt and portal access—even offline in dead zones.
A workspace where teams use existing plans and models—or create a model when none exists.
The operational layer coordinating models and data between Field, Studio and third parties.
The same evidence flow across different sites
Gewerkton is in beta now. Public beta is planned for fall 2026.
The platform is being built for global construction markets, but its origin is distinctly German. Its deepest commercial integration is with German systems and standards, including GAEB, REB, XRechnung and DATEV. Around that foundation sits a broader international proposition: 27 content languages, selectable AI providers across the EU, the US and Asia, including mainland China, and a choice between EU cloud hosting and deployment on a company’s own infrastructure.
Behind the scope is an unusually compact operation. Gewerkton is built by a solo founder directing a fleet of coding agents using Codex and Claude. In one night, that fleet shipped 21 software packages, verified with negative controls and mutation tests. The scale of the agent-assisted work is part of the story, but the platform itself remains focused on a familiar physical setting: people working on scaffolds, in tunnels, across wind farms, inside data centres and on housing projects.
Gewerkton is in beta now. A public beta is planned for fall 2026.
Tonfarb 136GB Digital Voice Recorder with Playback,9775 Hours Audio Record
- High-Quality PCM Recording: Supports 1536 kbps HD audio with noise reduction
- Large Storage Capacity: 136GB total storage with 9775 hours recording
- Extended Battery Life: Up to 68 hours continuous recording
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
A voice-first record of what happened
The idea behind Gewerkton is captured in its marketing line: “On site, what counts is what’s proven.” Rather than treating documentation as a separate office task, the platform begins with capture in the field. A spoken account can become evidence connected to a photograph, a deadline and a signature.
That approach is intended to reduce the distance between an event and its record. A defect can be captured with a photo and deadline. A daywork report can be dictated. At handover, a signature can be added on the device already in use. For instructions and change orders, the original audio can remain connected to the resulting evidence.
The platform is organised as one branded house with three product lines: Field, Studio and Cloud. Each deals with a different part of the same flow, from site capture to browser-based plans and models, then onward to operational and data coordination.
Field: dictation where the work happens
Gewerkton Field is the voice-first construction site app. It covers dictation to evidence, defects, daywork reports, takt and portal access. Takt refers to organising work around a planned rhythm, helping site teams follow the sequence in which activities are expected to move through a project.
Its role becomes clearest across the deployment fields Gewerkton identifies. On wind farms and renewable-energy projects, sites may be distributed, crews may rotate and field acceptance may happen far from dependable connectivity. Field supports offline capture in dead zones, allowing site information to be recorded where the work is taking place.
In housing and building construction, the emphasis is on practical records: defects with photographs and deadlines, dictated daywork reports and signatures on a device at handover. For infrastructure and tunnel projects, where work can run for long periods and involve many change orders, instructions can be backed by their original audio.
Data centres and industrial plants present another pattern. Many trades may be working in parallel under tight deadlines. Meeting decisions can become task lists sorted by trade, giving the spoken outcome a structured form for the teams involved.
The platform’s vocabulary is technical only where the work itself demands it. At heart, Field is about capturing an account without requiring the person on site to stop and reconstruct it later. Voice is the entry point; evidence is the intended result.
Studio: plans, models and a browser workspace
Gewerkton Studio is the browser workspace for plans and models. Where a model already exists, teams can work with it in the browser. Where no model exists, the site team can create one there.
This matters because the platform is not limited to projects that arrive with a complete digital model. Studio provides a place for the site team to establish that working representation rather than treating its absence as a stopping point. The model and plan environment then sits alongside information captured through Field.
For general readers, the distinction is straightforward. Field deals with what is spoken, photographed and recorded at the place of work. Studio provides the browser-based space in which plans and models can be used or created. The two are separate product lines, but they belong to the same evidence and coordination process.
Cloud: coordination across systems and regions
Gewerkton Cloud handles operations and the coordination of models and data between Field, Studio and third parties. It is the connecting layer in the three-part structure.
The platform’s regional choices are especially relevant for cross-border teams. Gewerkton supports 27 content languages, allowing EU, US and APAC participants to work on the same project in their own languages while the original evidence remains unambiguous. For projects in Asia, the stated use cases include Chinese, Korean and Vietnamese crews, with multilingual handling from capture through to report.
Data location is also presented as a choice. Organisations can use an EU cloud or their own infrastructure. Gewerkton’s bring-your-own-AI model extends that control: it supports 13 AI providers, lets users bring their own keys and makes the region selectable across the EU, the US and Asia, including mainland China. The aim is to avoid vendor lock-in while allowing provider and residency decisions to reflect the project’s region.
A solo founder and an agent fleet
The platform’s development model gives the launch a human centre. One founder is directing an agent fleet rather than presenting Gewerkton as the output of a conventional large software organisation. Codex and Claude are the named coding agents, and the reported one-night delivery of 21 software packages offers a concrete measure of that workflow.
Those packages were checked with negative controls and mutation tests. In simple terms, these methods are intended to show whether tests can detect failures and deliberately introduced changes, rather than merely confirming that software passes under expected conditions. The detail is significant because speed alone says little about whether a package has been meaningfully tested.
The same combination of compact authorship and broad output appears in the public-facing material. Gewerkton’s marketing site is available in 27 languages. It uses zero trackers, has no cookie banner and is built with a fully egress-free architecture. Its media bank contains more than 51 self-produced clips and posters.
That breadth does not change the beta status. Gewerkton is still in beta, with the public beta planned for fall 2026. The current story is therefore one of a platform taking shape, not a finished product presented as settled.
One workflow across very different sites
The deployment fields show how the same core workflow is meant to travel. Wind farms and renewable projects need capture across distributed locations and unreliable connections. Data centres and industrial plants involve many trades moving at once. Housing projects require photographs, deadlines, daywork reports and handover signatures. Infrastructure and tunnels add long durations, repeated changes and the need to retain the original audio behind instructions.
Cross-border work adds language and residency choices. Teams in Europe, the US and APAC can contribute in their own languages, while projects involving Chinese, Korean and Vietnamese crews can remain multilingual from initial capture to final report. Provider region and infrastructure remain selectable rather than being tied to a single AI vendor or hosting route.
Across those settings, Gewerkton’s central proposition stays narrow and understandable: speak at the point of work, connect the account to evidence, and carry that record through Field, Studio and Cloud. It is an ambitious global frame built around a simple site-level action—and, for now, a beta platform being assembled by one founder and a fleet of agents.
