Edmund AI: A Digital Colleague Who Talks to Machines
44 ǀ POSITIV TECHNOLOGY
TECHNOLOGY | TECH+
What does working with Edmund look like in practice?
He is basicly digital maintenance worker who ‘knows’
the entire production line – from documentation
and PLC programmes to repair history and employees
know-how. In practice, it tackles typical situations where
a line has stopped and a technician urgently needs
to identify what went wrong and how to resolve it.
Today technicians often rely on trial and error, search
through thousands of pages of manuals, or phone
colleagues for advice, Edmund instantly connects
all available sources. It extracts only the relevant
information and highlights the exact wiring diagram,
programme section, or documentation page that matters
for the issue at hand. The technician communicates with
it simply via chat, much like ChatGPT – but Edmund
focuses exclusively on technical problems.
A typical scenario is that the technician describes
an error or symptom, Edmund searches for possible
causes, and then offers step-by-step recommended
procedures. Thanks to its memory, it also remembers
previous interventions, so if the fault occurs again,
it can solve it even faster. Early experience shows that
with Edmund, problems can be resolved up to ten times
faster than with traditional approaches.
Beyond reactive troubleshooting, we are also developing
agents capable of acting autonomously to prevent
issues before they occur. And to make the transition
to such digitalisation easier, we offer companies data
health and data cleaning services – preparing their data
not only for Edmund, but also for future AI applications.
What makes Edmund groundbreaking? How is it
different from standard documentation search
or typical chatbots?
It brings a real-world implementation of AI directly into
production and maintenance – environments where
such technologies have so far been almost unused. While
smart tools are already standard in offices or finance
departments, this is still something new in factory halls.
Unlike universal models, Edmund is specialised:
it understands specific technologies, processes,
and machines. Today it can work with virtually any
documents and data generated in manufacturing – from
PDF manuals, datasheets, and repair procedures, through
PLC programmes, to database records such as spare
parts inventories, repair histories, or process logs from
production. Its greatest value lies in connecting these
sources and presenting them in context. A technician
doesn’t just see a wiring diagram or a line of PLC code,
but a complete, contextualised solution to the problem.
To achieve this, Edmund uses dedicated agents:
• The Documentation Agent finds the exact diagram
or passage in a thousand-page manual.
• PLC Agent analyses the control programme
and explains its logic.
• SQL Data Agent makes databases accessible – for
example, the spare parts inventory or repair history.
On top of that are features such as a maintenance
log, that records interventions and know-how,
and integration with the customer’s maintenance
systems. Practical tools include Suggestions – pre-set
queries for frequently recurring tasks – and Scheduling
Agents, which automate regular checks or reports.
All these agents collaborate, which is crucial for users:
rather than receiving isolated pieces of information, they
receive a connected overview of the problem along
with a proposed solution. In practice, technicians most
often rely on the documentation and database agents,
as this is where the majority of the information is stored.
Beyond Edmund itself, we also provide the data health
and data cleaning service, which can process massive
volumes of data and deliver a clear overview of their
quality – how many are duplicates, how many are
already digitised, and how well prepared they are for
further use in digitalisation and AI.
What do you see as Edmund’s competitive advantage
compared to other AI solutions in industry?
Edmund’s strength lies in its breadth and ability
to integrate. Most AI solutions in industry focus on a
narrow problem – for example, predictive maintenance
or data analysis from a single system. Edmund,
by contrast, can connect data across documentation,
databases, and specialised systems, and transform
them into one coherent picture. This gives technicians
and managers a comprehensive overview and enables
faster, more accurate decision-making – something
competitors do not yet offer on such a broad scale.
Thank you for the interview.
Edmund AI company was founded in 2023 and stand
behind the product of the same name – a digital
industrial assistant that is far more than just another
chatbot. It helps technicians and maintenance sta
resolve issues faster and more eciently, while
working with it is as simple as with any other LLM
model. No specialised IT knowledge is required – just
ask, and Edmund will nd the answer.
Experience shows that with Edmund,
problems can be resolved up to ten mes
faster than with tradional approaches.