Elea Ai is chasing a health care production opportunity by targeting old systems of pathology laboratories
VC financing was in artificial intelligence tools for health care It is expected to reach 11 billion dollars last year – A major figure who speaks to a large -scale conviction that artificial intelligence will prove to be transforming into the critical sector.
Many startups that apply artificial intelligence in the field of health care seek to pay efficiency by automating some administration that revolve around patient care. It is based in Hamburg Elia This template is widely suitable, but it begins with a relatively specialized and deprived prestige-pathology laboratories, whose work requires analysis of patients ’samples of the disease-from where you think it will be able to expand the scope of the workflow system in which AI is working in which AI works to enhance productivity of laboratories to achieve global influence. Including by planting its approach, which focuses on the workflow to accelerate other health care departments as well.
Elea’s initial AI is designed to fix how doctors and other laboratory employees work. It is a complete alternative to old information systems and other working methods (such as Microsoft Office to write reports)-converting the workflow to the “artificial intelligence operating system” that publishes copies of speech to text and other forms of automation to reducing the time it takes to diagnose.
After about half a year of working with its first users, Elijah says its system has been able to reduce the time it takes to produce about half of their reports to just two days.
Step -step automation
The CEO of Elea and co -founder of Dr. Christophe Schroeder says: The process of manual work in step by step means that there is a good room to enhance productivity through the application of artificial intelligence. “We turn this everywhere – and all the steps are more automatically … [Doctors] Talk to Elea, MTAS [medical technical assistants] He spoke to Eli, tell them what they see, what they want to do. “
“Elea is the agent, carrying out all tasks in the system and printing things – the slides, for example, staining and all these things – so that [tasks] Go so much, much faster, smoother. “
“It is not really more than anything, it replaces the entire infrastructure,” he adds from the program based on the group of casuals that they want to replace the laboratory LACY systems and the most wonderful ways of work, using separate applications to carry out different tasks. The idea of OS AI is to be able to organize everything.
Startup based on a different Language models (LLMS) by accurate control with specialized information and data to enable the basic capabilities in the context of the pathology laboratory. Bake the platform to speak to the text to copy the audio notes of the employees-and also “text to the structure”; In the sense that the system can convert these written audio notes into an active direction that operates the procedures of the artificial intelligence agent, which can include sending instructions to the LAB group to maintain the workflow.
ELEA also plans to develop its own foundation model for the analysis of the chip images, for each of the sheruder, as it is pushing towards the development of diagnostic capabilities as well. But at the present time, it focuses on expanding its initial width.
The startup stadium indicates to LABS that what can take from two to three weeks can be achieved using traditional operations within hours or days because the integrated system is able to stack productivity gains and gains by replacing things like the back and chosen that can surround manual printing about reports, where human error and other functional butterflies that can be associated with many Turkish.
The system can be accessed by laboratory employees through the iPad, Mac application or web application-offers a variety of touch points to suit different types of users.
The company was established in early 2024 and was launched with its first laboratory in October after spending some time in the ghost working on his idea in 2023, for every Schroeder, who has a background in applying artificial intelligence for independent leadership projects in Bush, Luminar and Mercedes.
Another co-founder brings Dr. Sebastian Casu-CMO at the start-up-background, after spent more than a decade in intensive care, anesthesia, and through emergency departments, as well as being a former medical director in a large hospital chain.
To date, Elijah has signed a partnership with a group of major German hospitals (not yet revealing any of them) that she says is about 70,000 cases annually. So the system has hundreds of users so far.
More customers are scheduled to launch “soon” – and Schroeder also says he is looking for international expansion, taking into account the entry of the United States market.
Seed support
The starting start for the first time is detected by a 4 million euros that I collected last year – led by Fly Ventures and Giant Ventures – used to build its engineering team and get the product in the hands of the first laboratories.
This number is a very small amount for the above billions of funds that fly around the area annually. But Schröder says that the startups of artificial intelligence do not need armies of engineers and hundreds of millions to succeed – it’s the most intelligent application of resources that you have, as suggested. In the context of this health care, this means following an approach focusing on the department and the maturity of the target use condition before moving to the following application area.
However, at the same time, it confirms that the team will look forward to raising a round (probably) of the A-Series probably this summer-saying that ELA will turn the equipment into active marketing to get more laboratories, rather than relying on the approach of the word what they started with.
Discussing their approach to the competitive scene of artificial intelligence solutions in health care, tells us: “I think the big difference is that it is an immediate solution to a vertical integrated.”
“Many of the tools you see are the additional functions above the existing systems [such as EHR systems] … it’s something [users] You should do in addition to another tool, another user interface, something else does not really want people who do not want to work with digital devices, and therefore it is difficult, and it certainly limits the capabilities. “
“What we built instead is that we have already integrated it deeply into our laboratory information system – or we call it the pathology operating system – which in the end means that the user does not have to use a different user interface, they do not have to use a different tool. He only speaks with ELA, says what he sees, says what you want to do, and says what is supposed to do ELA in the system.”
“You also do not need spinning from engineers anymore – you need tens, twenty, really, really good,” he says. “We have almost twenty engineers on the team … and they can do amazing things.”
“The fastest growing companies that you see these days, they do not have hundreds of engineers – they have twenty -one experts, and these men can build amazing things. This is the philosophy that we also, and for this reason we do not really need to raise – at least at first – hundreds of millions.”
“It is definitely the transformation of a model … how to build companies.”
Limbing the workflow of the workflow
The choice of starting with pathology laboratories was a strategic option for ELEA not only the one -handable market worth billions of dollars, for each of theherrrh, but it determines the space of pathology as “very global” – with international laboratory companies and suppliers who suffer from the ability to expand its program as a service theater – especially compared to the most courageous situation on hospitals.
“For us, it’s very interesting because you can build one application and already expand its scope with that – from Germany to the United Kingdom, and the United States,” he suggested. “Everyone thinks at the same time, and they behave the same, and it has the same progress. [and other languages like Spanish] … so it opens a lot of different opportunities. “
It also refers to the pathology laboratories as “one of the fastest growing areas in medicine” – noting that developments in medical science, such as high molecular diseases and DNA sequence, creates a demand for more types of analysis and more frequency of analyzes. All this means more work for laboratories – and more pressure on laboratories to be more productive.
Once ELA ripens if the laboratory is used, he says they may look forward to moving to areas where artificial intelligence is usually applied in the field of health care – such as hospital supporters to capture patients’ reactions – but any other applications they develop will also have a strict focus on workflow.
He says: “What we want to bring is the mindset of this work flow, where everything is dealt with like the workflow, and in the end, there is a report – and this report must be sent,” adding that in the context of the hospital they do not want to enter the diagnosis but “they will really focus on running the workflow.”
Image processing is another field that Elea is interested in other future healthcare applications – such as accelerating data analysis of radiology.
Challenges
What about accuracy? Health care is a very sensitive use, so any errors in these copies of AI – for example, relate to a biopsy that verifies the cancerous tissues – can lead to severe consequences if there is an inconsistency between what the human doctor says and what Elea hears and his reports to other decision makers in the patient care chain.
Chröder currently says they are accurately by looking at things like the number of letters that users change in reports that artificial intelligence serves. Nowadays, he says there are between 5 % to 10 % of cases where some manual reactions of these mechanical reports may indicate a mistake. (Although it also indicates that doctors may need changes for other reasons – they say they are working to “reduce” the percentage in which manual interventions occur.)
Ultimately, he argues that the Pax stops with doctors and other employees who are asked to review and agree to artificial intelligence outputs – indicating that Elea’s workflow is not really different from the old processes that were designed to replace them (where, for example, the audio memorandum is written for the ring.
Automation can lead to an increase in the volume of productivity, which can be pressure on checks that human employees must deal with with much larger data and reports for review more than they are used to.
On this, Schroeder agrees that there may be risks. But he says they are Benia in the “Safety Network” feature where Amnesty International can try to discover possible problems – using claims to encourage the doctor to look again. “We call her a second pair of eyes,” as he notes: “Where we evaluate the previous results reports with what [the doctor] He said now and give him comments and suggestions. “
The patient’s secrecy may be another concern associated with Agency AI depends on the treatment -based treatment group (as does ELA), instead of the remaining local data and under laboratory control. On this, Schröder claims that the startup has been resolved to penetrate “data privacy” by separating the patient’s identities from diagnostic outputs – so it mainly depends on the pseudonym to comply with data protection.
“It is always unknown along the way – every step does one thing – and we combine the data on the device that the doctor sees,” he says. “So we have mainly false identifiers that we use in all our treatment steps – which are temporary, then deleted – but for the time the doctor looks at the patient, it is combined on the device for him.”
“We work with servers in Europe, and we guarantee that everything is compatible with the privacy of data,” as it tells us. “Our main customer is a series of hospitals owned by the public – called the critical infrastructure in Germany. We needed to make sure that from the point of view of data privacy, everything is safe. They gave us the thumb.”
“In the end, we are likely to overlook what to do. However, as you know, it is always best to be on the safe side – especially if you deal with medical data.”
