CYKEL AI 

‍BROKER NOTE BY JONATHAN PLANT

Wednesday 4th December 2024

‍

We have always had a significant Lucy in the world…from Lucy who was an early australopithecine skeleton dated to about 3.2 million years ago with a small skull akin to that of non-hominin apes, to Scarlett Johanson’s portrayal of Lucy who is given high-tech drugs to take her mind to the edge of human capabilities…Cykel’s Lucy is set to be profoundly evolutionary for them also.

Cykel’s latest update¹ is a key milestone in their progress and makes them even more relevant in the AI space. The development of scalable, autonomous digital agents (life-like reasoners versus existing regurgitating Bots essentially), is a key differentiator and why they should see take up in a world which is thinking why do I need a bunch of tools I can’t use or program and won’t see the benefit from…this new iteration embeds into the workflow of a company and literally does work for the client.

“Lucy represents more than a technological advancement – she is a fully autonomous digital worker capable of executing end-to-end recruitment processes,” said Ewan Collinge, CEO & Founder of Cykel AI. “Our vision of the future workplace is one where digital workers handle routine operations while humans focus on strategic decisions. With Lucy and our pipeline of specialised digital workers, we are among the first movers in an emerging category that is expected to transform multiple industries over the next twelve to eighteen months”

The company’s ‘work-not-software’ model pioneers a new approach where organisations purchase completed work rather than tools, marking a fundamental shift in how businesses consume AI services.

‍

“Hi there,    I’m Lucy…

…I’m your 24/7 AI recruitment partner. I help talent teams reduce time-to-hire by automating sourcing, screening and admin tasks.”

That is how you will first meet Lucy but What is she, why and how?

Lucy is a recruitment-focuses automated digital agent, which can scale infinitely, with one Lucy able to handle the work of multiple human recruiters all day and every day…

automating the most time-consuming parts of recruiting: sourcing candidates, screening resumes at scale, writing job descriptions, and managing candidate communications. Lucy handles hundreds of applications simultaneously, letting the client focus on building relationships and making great hires.

• For the client it is all about return on investment ROI with most customers seeing a 5-10x return on investment through time savings and improved hiring outcomes²

• Customers typically see up to a 30-50% reduction in time-to-hire, 8+ hours saved per recruiter per week, 90% reduction in resume screening time, and improved candidate quality through consistent screening^2a. Lucy starts delivering value from day one, with most teams seeing significant improvements within the first month.

‍

Are clients concerned about losing control of their processes?

They shouldn’t be as full control is maintained over screening criteria and Lucy is just a flat-out digital worker augmenting the existing team not replacing them…it is designed to help you scale up the business efficiently…they are hiring her basically³

Connectivity is almost seamless as Lucy will integrate with most major Applicant Tracking Systems (ATS) and has industry leading security protocols to keep the client and their data safe. In recruitment parlance this maintains a single source of truth (SSOT) into the ATS…basically allowing the client to have increased confidence over their process being robust and accurate.

‍

Why AI and why Cykel?

‍

Plain and simple… the Sweet spot has arrived.

They have done their due diligence, built the models and processes.

If you want to be in the game now clients are looking for autonomous digital agents and potentially swarms of them that can weave through the business lines. Is having a face a gimmick, style over substance? No, it is the key to proliferating, seeing mass take up, as the human touch crosses the divide. You are literally working with something that can react like a human. Digital workers differ from normal software tools as they learn on the job and improve over time like humans do.

You’re not just playing with an app or browser extension, stuck in the world of text processing. You actually have the ability to seamlessly integrate AI automation into your workflows.

• Clear trend to improving ROI by using AI, and Cykel’s in particular, as mentioned above.

• Cykel’s partnerships model has driven industry specific processes allowing digital agents like Lucy to reduce operational costs while increasing productivity, validating Cykel’s position as one of the leaders in deployable digital workforce solutions. With additional specialised digital workers in development, Cykel is set to expand its reach into more value added areas throughout 2025 and beyond, becoming ever more relevant in the space. This will place them into the world of Agent Swarms which will be another term that you will hear more of over 2025.

These partnerships have sequentially built upon the developments made through the workflow tools created so far…they haven’t reinvented the wheel or gone down rabbit holes, they have used tie-ups such as the one with Chaptr⁴ to drive the recruitment opportunity by building tools and relationships from bottom up, utilising experts, outsourcing with talent partners, to produce efficient solutions…this is technology proven in production not just a one in the box solution. In terms of Cykel as an investment they haven’t been wasting their time they have been going through multiple iterations of development to stay relevant into the next monetisation period. This sector is obviously new and expanding quickly. As we will show below the potential for revenue catch up is incredibly high.

• The Lucy concept itself is industry agnostic and Cykel already have plans to expand into niche areas. A generic sales agent is a saturated market, what is not is industry specific solutions…so expect developments in research analysis and nuanced sales areas.

• This concept has the ability to be multi layered and hand off tasks to other agents…vertical integration…creating a network of understanding in the client process that is hard to replicate…this creates the real first mover advantage for Cykel as once they are entwined into the client’s processes that is hard to replicate and should lead to not only low churn, if any, but create whole client upselling opportunities.

• Not just UK centric…you can see from their partnerships such as Simera⁶ in Boston, that the US is on the growth path and obviously the digital world is not enclosed by borders.

A crucial concern for clients is opening up their often customer sensitive data into what could be this autonomously controlled digital environment. Here Enterprise features (compliance, security, audit trails etc.) become even more valuable as digital workers handle sensitive tasks.

• Security is the key…if you are dealing with sensitive, legally compromising data you need a partner that is safe, has the security technology covered and has a regulated responsibility which Cykel does with it being listed on an exchange, the LSE, as a Public Company, and is also GDPR compliant²⁸ . There are a lot of start-ups looking to cash in but you need a partner that is there for the long run. Someone messing with your coding or even inexperienced internal attempts at re-creating the wheel are dangerous to company profitability and longevity.

Cykel maintains the highest standards of data security and compliance including full GDPR compliance, enterprise-grade encryption, secure data handling practices. Client data will never be used by Cykel or their LLM providers to train models. This gives clients confidence they have a one stop solution.

‍

Updates since last note

‍

Recent partnerships demonstrate Lucy’s ability to reduce operational costs while increasing productivity, validating Cykel’s position as a relevant player in deployable digital workforce solutions. With additional specialised digital workers in development Cykel is set to expand across other sectors and some of the crossover has been prepared for as you will see below.

Following on from the UK SHIFT partnership⁵ giving exposure to Fintech networks and the Chaptr Recruitment tie up⁴ there have been a number of press releases from Cykel.

• There is a clear concentration on the recruitment/HR space with another tie up with Simera in Boston, US, that is focused on the Latin America into North American market⁶

• Recruitment focused workflows and an ATS Applicant Tracking System have been developed with the Chaptr workflow developments revealing that 70% of the recruitment process can be automated⁷

• Development of a web application complements the browser extension versions of TaskOS and is clearly a defined path towards the Lucy solution⁸

• The Babelbond tie up is interesting as they are focused in AI translation of fintech, real-time translation of financial communications…this is an obvious next step for Cykel and the automated agents to be language agnostic and is becoming an expected role for AI solutions⁹

• Cykel will be looking to roll out automated digital agents into other sectors…the Legal-Pythia tie up is an obvious area for development since there is ahigh percentage of repetitive tasks and search work required with the Legal world looking to free up more time for critical thinking and value-added business¹⁰

• Vero Tech is a 3^rd party sales relationship¹¹ with Cykel looking to leverage their Leads and Scale for outreach campaigns…these have and will create large scale interest but has also shown that there are long lead-times in the sector…interest is high but follow through is delayed either from wanting to try first or the client’s data is just not ready to start…as we will show in the numbers section this has a double effect of backing up the revenue delivery time but gives it a higher confidence of large lumpy growth as once clients commit they don’t want to be looking for another partner nor running numerous different apps…churn should be low.

Overall, it is clear to see the direction of travel here with breakthrough in the recruitment sector, languages integrated and further scalable sector rollouts to follow.

‍

Sector

‍

There is a pivotal shift toward AI systems that can actively engage with computer interfaces rather than just process text and images.

OpenAI CEO Sam Altman said “we will have better and better models,” but “I think the thing that will feel like the next giant breakthrough will be agents.” At an OpenAI press event ahead of the company’s annual Dev Day last month, chief product officer Kevin Weil said: “I think 2025 is going to be the year that agentic systems finally hit the mainstream.”¹²

This confirms Cykel’s view that 2025 will be the year of the digital worker; they see a major AI sector development being the larger player models plateauing; now needing to sell the AI transformation. Exponential infrastructure growth is beginning to morph into ground up applications…This will answer the whole monetising argument for the sector. Cykel is positioned to be a market leader alongside companies like 11X.AI and TEZI.AI.

Nvidia results suggest the market growth rate is still huge even if the spend growth rate may be slowing, but it is not declining. There is a difference. The billion-dollar opportunity now has to literally be generated. Advertising budgets are huge with the mag7 needing to create the buzz to justify spending capex levels. This will all help focus eyes on companies like Cykel.

Due to various anti-trust issues with Europe and Apple their AI agent is likely to see delays in competing in the UK/EU environment; although this may become a negotiation point in the Trump administrations Tariff war.

Data problems at client level has been delaying implementation more than a desire to take up AI…the Cykel approach to implement a worker into the system, focusing on process over data point, should resonate with CIOs. Success will come to self-learning and responsive solutions rather than straight pass-through text handlers.

The challenge for CIOs is preventing these tools from creating a whole new world of Shadow IT. If non-developers start using AI tools for internet-facing systems that handle customer data, this could create serious security risks. That is why it is critical that the tools are used as an aid and with the support of skilled developers who understand industry regulations, coding standards and the security risks associated with the project at hand³.

Lucy is orchestration level, which is the coordination and management of multiple computer systems, applications and/or services, stringing together multiple tasks in order to execute a larger workflow or process. These processes can consist of multiple tasks that are automated and can involve multiple systems.

Here Cykel is one of only a few players at this level of complexity.

A phrase you will come across searching the AI universe is…”people want a one stop shop” and Cykel are moving down that route…vertical integration, security, multiple languages etc.

Einstein AI for America…access to a genius for teaching, learning, processing that can also code, learn etc…non-human AI Scientists, specialists very close according to Eric Schmidt, former CEO and chairman at Google and co-author of “Genesis” a book about the hopes and future of AI¹³.

Salesforce paid $20m a decade ago to purchase rights to use Einstein as the face of their AI platform but essentially, they are all about CRM (Client Relationship Management) plus Data plus AI plus Trust¹⁴…everything that Cykel has with Lucy too.

Lots of people are trying to enter but where are they on the technology curve?

There will now be an acceleration of solutions and Cykel’s is extremely relevant.

As we have said previously keep an eye on the top layers in the whole AI space…this concept is just beginning so forget linear growth – AI is poised for a hockey stick trajectory¹⁵.

Imagine every sector, from healthcare to finance, infused with superhuman intelligence.

Palantir (NYSE:PLTR), the Messi of AI, has had an incredible month post results and through a Trump victory…they stand to benefit from large Government contracts amidst a productivity drive in the new administration. But they are more than a political play they too have recently released an Artificial Intelligence Platform (AIP) which has contributed to a “game changer quarter” with AIP monetization as a “major growth driver”. Goldman Sachs has recently called them an AI Growth story¹⁶ …well if they can grow from their starting point how about an exponential launchpad at Cykel?

Following Palantir Technologies Inc.’s Q3 earnings, the stock soared by 47% over the next week or so. The results saw the firm’s Quarterly commercial revenue jump by 54% as it indicated strong demand for its AI code evaluation platform.

Palantir management mentioned AI during its Q3 2024 earnings call:

“We’re witnessing the commoditization of cognition with the rapid advancement of AI models. Almost all investment in the AI space has been focused on supplying and improving these models. What will differentiate the AI haves from the have nots, is the ability to maximally leverage these models in production by capitalizing upon the rich context within the enterprise. This is Palantir’s focus. We see this in the results we’re delivering for our customers. Those who embrace quantified exceptionalism through AIP are able to take advantage of the commoditized cognition in a levered way to advance their differentiation. In this winner take all AI economy, the divide is widening between those who are leveraging AIP and those who are not. At a leading global insurance organization, AIP has helped automate key underwriting workflows, reducing the typical underwriting response time from over two weeks to 3 hours.”¹⁷

Does that sound familiar?

On top of that there is clear direction in Government’s engaging in the space with the French government offering again to Buy Atos Origins for its AI computing capabilities and keep their AI capabilities in-house as it were¹⁸. Cykel is not involved with the UK Government currently but this shows just how high-level interest in AI has become. We’ve also shown in a previous note how Sovereigns are increasingly investing in SaaS and AI so the proposed UK super pension fund by Rachel Reeves recently could well look at this space if they are looking to invest in future trends. I have not seen this discussed but is not a giant leap in the thought process.

‍

Valuations

‍

The biggest players in the infrastructure have been busy since our last update. Open AI has re-evaluated into being a for profit model (although details have not been released) and has come to market again for funding. The latest round raised $6.6B in new funding at a $157B post-money valuation to accelerate their progress; with chipmaker Nvidia, Microsoft and Japanese group SoftBank among its investors. Followed up swiftly by Softbank tendering $1.5bn for Open-AI employee stock, who have until Dec 24^th to decide, to try and increase its shareholding further from the $500m in the latest funding round¹⁹.  Much has been made of insiders selling in the tech space, cashing in for some development lifers, but it seems the investment frenzy is still much alive.

You don’t have to be a big player to be on the trending list just in the right space and reaching the right investors²⁰

11X.AI two raising series A and B rounds at the level we can expect Cykel to progress towards²¹

11X is a startup that develops AI-powered sales development bots…they have gone through two funding rounds raising $24m at a read through $90m valuation and recently $50m at a $350m valuation. This latter deal was based on c. $10m ARR (annual recurring revenues) which is roughly where we still anticipate Cykel being in two years’ time and implies 35x is not out of the question when we have previously been ultra conservative at 8.6xRevenues for genAI?

Hebbia, a startup that uses generative AI to search large documents and respond to large questions, has raised a $130 million Series B at a roughly $700 million valuation led by Andreessen Horowitz, with participation from Index Ventures, Google Ventures and Peter Thiel. They are used by asset managers to do due diligence and research investments…(this is an area that would be a good fit for Cykel to consider for its digital workforce expansion)… its funding demonstrates that 50x ARR is not unusual in these value added areas for AI startups, especially ones that have booked millions of profitable revenue early in their journey²²

Tezi.AI received $9m of Seed money to do the LLM (Large Language Model) side of recruitment that Cykel is in. So lots of potential investors in this recruitment space .with Cykel at the front end with the digital agent launched²³

‍

Numbers & Revenues

‍

Lucy is available to try for free for the hobbyist individual and pro individual level but both with limited usage. This may be an ideal way to trial for some smaller players but after the current trial period of a month finishes, we assume that January 2025 is the launchpad for revenue growth.

Cykel’s focus is on the Enterprise level where companies may have up to 20 users or more with 2k of credits/user/month. Originally, they wanted to charge for workflows in a customisation package way, for larger projects, but with the whole self-learning and vertical integration concept, which should see them capture most of that expected extra spend, then Enterprise is likely to dominate the participation level.

Once the industry specific agent is released the client will see specific development…this should keep take-up levels high and larger organisations arranging higher usage deals. There is an incentive to sign up for a discounted annual rate of $79/user/month versus $99 on a monthly contract. For the Enterprise set up I am assuming 15 out of 20 users per client to allow for rolling start up and a blended rate of $92 which is 66% monthly and 33% annual to allow for the nature of trialling and scaling plus an element of clients growing above 20 users.

To keep things relatively simple in the model I will assume there is 1 custom client at outset in Jan25 and grows in line with the user rate proposed per month in each year but that total workflows will only grow at 1% per month and spending $250 per month/ workflow. This should also capture some of the scaling likely if the enterprise growth moves towards the maximum 20 users.

There will be a few free users around as always, possibly trialling for their company, and many of these will need to use the pro level tool for testing. I have assumed 50 free users at Outset, 25 Pro, 100 Enterprise and the 1 Custom as mentioned above.

The pricing²⁴ is designed to encourage enterprise take up and sign up to annual recurring revenues as opposed to monthly.  As benefits are delivered you would expect the mix to move towards more annual Enterprise sign ups.

Given the networking they have done teaming up with the partnerships we have highlighted above, and actually developing this from within the recruitment industry, there should be a high degree of interest already from the sector even before the marketing post the Lucy announcement begins. You can’t watch business media without an AI advert currently so they are hitting at the right time.

Until we see actual numbers the growth set up we have should capture any transition between users and enterprises.

‍

Assumptions

‍

There is a platform effect whereby customers who take up Lucy are more likely to adopt future releases/digital workers due to familiarity and trust.

As these agent swarms grow they should increasingly deliver value for the client in adjacent workflows performing a complex combination of tasks which then translates into further spend. Higher value workflow agents such as the Research engine will be operating in a field (Private Equity, Venture Capital, Investment Banking, Consultants etc) where the incumbent humans are high cost so there is a likelihood the whole value chain is larger as Cykel expands.

So in all likelihood we could be underestimating the revenue growth rates once uptake begins as multiple layers and higher value sectors could be in play, but from an overall narrative point of view this shows that the delay to monetisation is potentially short lived. The more sophisticated product is competitive and likely to see mass adoption when it starts.

I was relatively comfortable with the original figures pushed back by a year, which is more than offset in the underperformance in share price, since the last update from the company. If anything, the processes and workflows have been refined, with a hands-on approach alongside clients. Those processes are built into Lucy now and give an extra confidence that the growth will be achieved.

Pushing the new metrics through the model with start-up Saas growth rates, you can see in the table below how that translates to revenues…it already suggests there could be upside to previous revenue projections.

‍

‍

Now this is revenue projection and ignores costs but looking at an excellent piece about the rule of 40 for Saas and AI²⁵, something we discussed in the last note, but showing signs of becoming the rule of 55 for the successful companies we can make some assumptions about the DCF model below.

‍

‍

We run the DCF for fully paid and also fully diluted including options and warrants outstanding. We assume 15% WACC and a perpetual growth rate of 5% which is in line with the current US GDP nominal rate and fx rate of £/$ 1.2671 as at Nov 28^th 2024.

This Provides for Revenue DCFs of £3.90 and £2.23 respectively but to have any relevance we need an assumption for Free Cash Flow FCF, The note²⁵ suggests that 23% for an operating CF for the successful genAI companies is a classic result. Applying this to the two revenue DCFs we derived provides for a FCF DCF value range of £0.897 – £0.5129.

From the readthrough valuations in the space we can see some similar valuations to give us confidence that if Cykel has hit the sweet spot their attraction will follow a similar path to 11X.AI and Habble that we showed above.

35x FY26 revenue estimates of $12.6m = £348m

Immediate progress towards the $3.13m FY25 estimate could open up confidence for a Habble type backing at 50x which would suggest Cykel being valued at £122m and confidence in that higher FY26 valuation.

A mature multiple of 8.6x for GenAI that we have conservatively used before suggests that as we progress towards FY27 where the monthly user growth rate may be slowing to less than 10% (7.9% in the model) then the multiple for that period gives a valuation of £260m.

Given the speed of advancement in the AI bottom-up market, the fact that Cykel have the right product for current trends and a product/price mix that encourages take up with strong loyalty and upselling features, plus a trend for investment in the winners we have confidence that it is reasonable to have a price target of 65p within the next 24 months. This will require progress to be made on revenues within 2025.

Further support for the genAI Global marketplace:

The staffing and recruitment market size is expected to reach US$ 2,031.34 billion ($2Trn.) by 2031 from US$ 757.56 billion in 2023. The market is estimated to record a CAGR of 13.1% from 2023 to 2031. The rising adoption of automation and AI-driven recruitment models is likely to remain key staffing and recruitment market trends²⁶

AI could contribute $15.7trn to the global economy by 2030² according to a PwC report with financial services being the main beneficiary.

Bloomberg Intelligence suggests below that genAI could be 12% of Total Tech spend by 2032 and be c. $1.3trn in revenue²⁷

‍

‍

Conclusion

‍

Why should you have confidence in Cykel?

Workflows and genAI appear to be the avenue of choice for the successful players in the sector just as 2025 is the year for adoption. Even though Cykel is small it is punching above its weight with its technology offering and is on the scene for take up. This will be relevant again for finding development growth via funding as we have shown with peers being backed heavily.

Trends have been for a lot of trial engagement before firms fully commit and has been the case for the large successful players but also why they now have large cash flow contracts. Success leads to loyalty and upscaling that sees cash flow and revenue profiles that will appear blob like in updates and form a step change in growth.

Look at the Messi of AI, Palantir. They have said this quarter was transformational for them on their AI platform launch…growing revenues at 54% driven by demand for its AI code evaluation¹⁶…that is 15.48% per month which is above any month we have in for the relative start-up who should be growing quicker!

This is why the lack of immediate revenue appearance is not a concern so far as it has increased the likelihood of Cykel delivering quickly, and in size, due to their continued front of class product development.

When this comes through it should have the exponential profile of an adoption curve model, the hockey stick curve¹⁵

If Cykel is going to breakthrough this is the time…in AI terms Lucy and her digital agents are at the cusp of what is possible within this space and Cykel are well placed to benefit from the humanisation of AI.

‍