Here is Joni’s latest company report on Digital Workforce.
We reiterate our target price of 4.7 euros and our ‘add’ recommendation for the share. Digital Workforce’s financial statement was well in line with our expectations regarding revenue, profit, and guidance. Additionally, the company commented that its sales pipeline remains strong and its quality has improved. We keep our forecasts largely unchanged and expect the company to grow significantly better than the IT services market in the coming years, with growth scaling well into profitability. The share’s valuation (2025e EV/EBIT 16x, P/E 18x, sum-of-the-parts 5.2 €) supports our positive view on the share. As the turnaround in earnings continues, the company’s very attractive long-term valuation picture can be relied upon more strongly.
I’m pondering how this small company’s systems operating in the healthcare sector remain competitive and relevant in a rapidly evolving market, given that larger players have significantly more resources, or are DW’s solutions unique and “difficult to copy”?
Quoted from a broader report: The company was founded in 2015 to originally help healthcare organizations with information work automation. Now, the service offering covers the entire automation lifecycle: planning and consulting, development and implementation, a cloud-based platform, support and maintenance, and continuous development.
Domestically, one might assume it’s a strong player, but considering this company’s longer-term prospects, how will it fare? A small company with small figures, yet possessing strong expertise in its niche, could DW be an acquisition target?
Perhaps these are a bit silly questions, and if they are too silly, there’s no need to answer, but I thank you in advance for any potential answers.
Good questions, here are my views on DWF as an investment and its risks. By the way, DWF comes from Digital Workforce’s ticker.
DWF does not itself manufacture products, but uses products from several different companies in business process automation and management (BPM).
DWF manages all products used in its cloud service (Outsmart), so the customer does not have to install the products themselves; DWF only provides the software to customers. The customer pays a regular fee.
This is convenient for customers and is apparently a partially scalable service for DWF, meaning new customer accounts bring more revenue than costs.
In addition, this brings stability and predictability to revenue and profit.
DWF is not committed to specific products. For example, if sema4.ai’s Agents are a poor solution, DWF can use, for example, UIPath’s AI Agents.
DWF’s competitive advantage is also that it focuses on specific operational areas, especially healthcare, possessing not only technical but also business expertise.
It is also already reasonably large.
In Finland, competitors include at least Alvoration Oy and Sisua Digital Oy. They are small, and Sisua is unprofitable.
The size of DWF’s orders is growing, and it is investing in the ability to manage increasingly large projects and initiatives.
AI Agents
Software robotics (RPA), which is reliable, is used in process automation. AI Agents can be adjusted by the end-customer themselves; they are more versatile, but their operation is not entirely predictable. Therefore, they cannot be used in critical applications.
Hybrid solutions combining RPA & AI Agents are likely to be sensible in the future.
sema4.ai collaboration involves risks and opportunities.
The risk is that sema4.ai is a startup and likely still an unprofitable company operating on venture capital, which could go bankrupt if it doesn’t acquire customers.
However, I believe that the development of sema4.ai’s agents will continue one way or another, despite a possible bankruptcy.
DWF could, of course, just switch to UIPath’s agents, for example, but investments in sema4.ai would be wasted.
Many large players are developing AI Agents, such as OpenAI Operator, but these are currently aimed at consumers (e.g., “book hotel rooms in city x”).
Major competitors in enterprise-level process automation include UIPath.
sema4.ai
sema4.ai can use several different language models (GPT 40…).
sema4.ai builds functionalities specifically for AI Agent management and the ability to integrate and process company data.
An example is versatile integration into various data sources (such as Snowflake).
Or an easy way to define what data from a document is processed, as the one-minute video below demonstrates:
sema4.ai & DWF
DWF recently hired coders for AI Agents. They can make extensions to sema4.ai’s functionality using sema4.ai’s programming libraries.
If successful, this could give DWF additional competitive advantages, and DWF would take a step towards becoming a technology company.
There are many technical possibilities here, and results are expected after the summer.
International expansion
This is difficult for a small company, but active operations have brought DWF’s name to prominence at least in the UK and through cooperation with a US healthcare system manufacturer.
I hope DWF is the acquirer and not acquired. Many players automating processes for customers are likely still quite small abroad, which is why DWF has growth opportunities.
Thanks @JukkaM for the clarifying article on the technology side!
AI agents seem to be the hottest thing right now, but for a layman like me, assessing DWF’s competitiveness is a challenging task, and videos like this don’t shed much more light on it. https://youtu.be/k5Ya3jKKQEk?si=m71zBrj8F4i2Heoi
In these situations, one has to seek support for the narrative from the numbers. I haven’t built my own Excel yet (and I’m considering whether it’s worth spending time on such things in these AI times), but I’ve been glancing at Inde’s charts and I’m shamelessly sharing them here from behind a paywall. However, it’s such a quiet thread and a little-traded stock that I think sharing these will do more good than harm.
Revenue seems to have turned to slight growth, but based on the numbers alone, it’s still difficult to assess how sustainable it is.
The share of recurring revenue has been trending upwards, which can be considered a good thing for profitability.
At least based on the costs, it would seem that some slight scalability is occurring.
Previous discussions would suggest that the market has not always had faith in the management and main owners. For example, when entering the stock exchange, the company ambitiously announced a target of 100 million in revenue by 2026. Since then, it has been modestly reduced by half, of which 40 million would be organic and 10 million acquired through acquisitions.
However, it now feels like there has been good buzz around the company lately.
Achieving the organic target would require robust annual growth of over 20%, yet the company only “estimates revenue to grow in 2025”. Targets are targets and guidance is guidance, but in my opinion, these are somewhat out of proportion.
However, for last year, in its guidance, the company “estimated full-year 2024 revenue to be higher,” and with such guidance, it achieved 9% growth. Of course, no one can truly predict these things, but the growth in the recurring revenue share and a “good sales pipeline,” as mentioned in the interview, could be expected to bring that growth.
Perhaps “growing” then means something more than 9%. Inde predicts it to mean ~14% growth for the next couple of years.
Inde’s forecast only includes organic growth, which thus falls short of the target, but even with those figures, the company would be quite attractively valued.
I have been building a small research position in the hope that it would motivate me to study the company and the industry more, and perhaps even type out my own Excel. More learning and understanding, as well as evidence from the company, are still needed before I dare to increase the position to a larger weight.
I’ve only been following this myself for 4 months and don’t know the history other than what I’ve read in some old reports.
From those 2022 videos, one could only conclude that some kind of turnaround has occurred after 2023, not just in numbers but also in personnel.
Jussi Vasama was appointed in spring 2023. Profitability, in particular, has started to improve after Jussi’s arrival.
For me, alongside the numbers, a good product/service and competent leadership are key. Technology companies should have strong technical leadership; this is invariably the case with US companies, but often not in Finland.
Karli Kalpala was appointed Head of Strategic Change and UK Area Lead in February 2024.
I have read quite a few of Karli’s articles and listened to an interview on the Signaali podcast, and based on those, he seems like a suitable person for a technical company’s strategy director.
Additionally, Antti Karjalainen as CTO is already excellent due to his merits and will enable the necessary steps to be taken, if desired, also towards in-house development.
So, quite a turmoil has taken place in the personnel, perhaps from a somewhat dusty state to a more active direction.
The company also, as I understand it, laid off personnel who did not fit the strategy.
International marketing is also actively present nowadays.
The €50M revenue target for 2026 seems unrealistic, as no clear organic growth was guided for 2025, which you also referred to. An impossible growth would be required for 2026, unless 2025 growth surprises.
Of course, many others, like Tamtron, have a somewhat similar sky-high target.
The poor economic situation has created a need, especially in the social and healthcare sector, to improve productivity, and DWF seems to be getting involved in that.
DWF should not, of course, spread itself too thin, but Germany could be an interesting market. The economy is in a bad state, and in Germany, there are, as I understand it, few SaaS services in use. The USA might be a competitive market, but time will tell.
In my opinion, the renewal of leadership has brought sharpness to the strategy, and the expansion of AI use to streamline company processes could provide an additional boost. Of course, there are also many risks in that AI Agent scenario, but fortunately, DWF has experience-based expertise.
I have a relatively small position and will add more as my confidence grows further.
I also bought a bit of UIPath, which makes products for RPA and AI Agents.
So now there’s a service provider DWF and a product manufacturer UIPath.
However, still relatively small positions.
As a product manufacturer, UIPath scales better than DWF, but like DWF, it is on its way towards positive earnings growth, where valuations are rapidly decreasing.
One reason for buying these stocks is the trend of how AI models are currently being developed.
The applications of AI Agents are currently limited by their reliability in critical tasks. Of course, they have several advantages that RPA cannot achieve.
However, currently, for several reasons, reasoning capabilities are specifically being developed in language models. For example, in Meta’s Llama 4, reasoning capabilities have been improved specifically considering the needs of AI Agents. META Chief says Llama 4 will power AI-Agents
So by the end of 2025, the reasoning capabilities of AI Agents could be much better and they could be more widely applicable.
This trend is also accelerated by the fact that over the past 3 years, many solutions have shifted from combinations of code and AI models towards solutions that rely solely on AI models.
This is the case, for example, in autonomous cars (e.g., Tesla FSD), robots, and AI Agents.
DWF will benefit if it is involved at the beginning of this trend.
The Finnish (SOTE) market is not sufficient in the long run, so the big question is how DWF will get business abroad.
Clear, concrete benefits obtained immediately after deployment with the help of DWF’s robotics. A clear win-win for both healthcare and patients.
“” In the emergency department of the Pirkanmaa wellbeing services county, a robotics solution was implemented at the beginning of February, which automatically identifies frequent attenders among patients, and enables their more efficient referral to primary healthcare for assessment of care needs. The solution improves the quality of care and enhances the utilization of emergency department resources.
..
During the first month, nearly 7,000 patients have gone through the process, of whom 10% have been identified as frequent attenders. For 6% of them, a contact request has been made to primary healthcare with the patient’s permission. For these patients, a follow-up contact is arranged within two weeks, during which their need for care is assessed and they can be directed to appropriate treatment.
“”
Outsmart is just a cloud environment from which DWF provides all services used in processes (sema4.ai, UIPath…) with a regular monthly fee. The customer doesn’t need to install anything. New customer fees are higher than DWF’s costs, so it scales somewhat.
That A2A is for AI Agent information exchange. So, sema4.ai and UIPath, used by DWF, will probably support it in the future. Their tools will likely include the ability to define the data structure (schema) for communication between different Agents.
This thus improves the building of multi-AI Agent environments and their communication.
I asked ChatGPT to draw the images
So, that Business Process corresponds, for example, to people from whom confirmation for a process is requested, company systems, or process-triggering factors (such as a new order..)
Above are the schemas for communication between different agents
DWF has made videos that talk about various current AI topics. The last topic was “Shopify CEO’s ‘Leaked’ AI Memo, OpenAI’s Memory Update, and Google’s NotebookLM”
In my opinion, that concept should be changed. General information about new LLM models and GenAI topics is already heard here and there.
In my opinion, DWF should talk about its own business, just like NVIDIA, for example.
For example, AI Agents are not suitable for all types of tasks, which is why RPA is likely used together with AI Agents.
DWF should explain what types of use cases can currently be done with RPA and AI Agents.
On the other hand, as the reasoning of LLMs used by AI Agents develops, those cases could be explored.
Also, this A2A and, for example, Anthropic’s MCP protocol (Model Context Protocol) could be explained at a general level, spiced with good images.
For these presentations, one person, Niko, who has made clear slides, would be sufficient. The video doesn’t need to be longer than 10-15 minutes.
Currently, in the videos, two people, Niko and Kristiina, are chatting.
Edit This seemingly technical addition might be important for the commercial utilization of DWF’s AI Agent usage.
The use cases for AI Agents are limited by their uncertain operation, due to the LLM they use.
The precise structures of A2A for AI Agent communication might increase the probability of AI Agents’ reasoning. Furthermore, in the development of LLMs, A2A can be better taken into account, which also increases the probabilities of AI Agents’ reasoning.
This expands the use cases for AI Agents and thereby grows DWF’s market.
Digital Workforce will publish its business review on Friday sometime in the morning, and here are Joni’s comments regarding it.
We expect last year’s strong revenue development to have continued in Q1. We predict growth to scale into profitability, but at the same time, investments will limit the improvement. In addition to operational development, the company’s comments on Finland and growth markets, especially the USA, are central.
Below are Joni’s quick comments on DW’s Q1 results.
Digital Workforce, a pioneer in automation utilizing software robotics, published a Q1 business review this morning that was weaker than our expectations. Revenue decreased and fell short of our expectations. Driven by the decline in revenue, the result was negative and below our forecast. However, the company implemented corrective measures and received new orders, which should support development already in Q2. The company reiterated its guidance and expects revenue and adjusted EBITDA to grow in 2025.
The figures were indeed weak compared to what the guidance, the CEO’s interview in connection with the financial statements, and the recently set financial targets indicated. Could you please challenge @Joni_Gronqvist a bit in the interview now, asking how the management believes they can achieve such targets when the numbers are like this?
“The financial results for the first quarter were weaker than anticipated. This was influenced by significant uncertainties in the international economy and trade, as well as the general weakness of the IT services market.” I would also be interested in how these significant uncertainties in the international economy and trade specifically affected Digital Workforce when they didn’t have time to affect others in the first quarter? This seems like a very strange explanation, and the IT services market has been weak for a longer time, so that couldn’t have come as a surprise to anyone anymore.
Ambitious financial targets, updated dividend policy, share buybacks, talk of a strong sales pipeline, AI, AI, AI… Next, probably some Bitcoin strategy?
Here is an interview with DW’s CEO Jussi Vasama, interviewed by Joni, of course.
Topics:
00:00 Introduction
00:08 Q1 Highlights
01:07 Background of the decline in expert services
02:40 Delayed projects
04:07 Recurring services decreased from the previous quarter
04:59 Sales development by country
07:45 Sales pipeline
09:03 Profitability below expectations
10:34 Cost savings
10:58 AI has developed really fast
13:14 How to turn the development around?
14:03 M&A market
Digital Workforce has received some good news in recent days, which should lead to the company’s revenue returning to growth. On the selling side, for several months, there have been practically no others than Danske and Aktia funds, which are apparently dumping the stock without caring about the price. When will market interest awaken?
