Promotion: Capital Invested is at Risk. Non-Independent Research / Marketing Communication. Attention is drawn to the disclaimers and risk warnings at the end of this document
BROKER NOTE BY JONATHAN PLANT
Wednesday 4th December 2024
In the Wild West shoot-out of the AI agent world it’s perhaps apt that the action is going to take place in the saloon, only in this one we will meet Kondor and Ora at the Sundae-Bar. For those not akin to what that is a sundae bar is a fun, interactive dessert experience where guests can customize their own ice cream sundaes. It typically features a variety of ice cream flavours, toppings like sauces, fruits, and sprinkles, allowing for creative and personalized dessert creations. The analogy reflects the ability to customise and personalise the Agent AI experience. The bar in the professional sense implies monetisation of the creation.
Where are we in the agent AI space?
The pace of change has caught out even those inside the market. New research is forever increasing the size of the opportunity in AI and Agentic AI
AI market is booming and according to Statista Market Insights is estimated to grow to $826.7 billion by 2030 (CAGR of 27.67%)note1 whilst AI Agents market is expected to grow from $5.1 billion (2024) to $47.1 billion (2030)note2. 85% of Enterprises are expected to use agent AI in 2025note1a.
Computing power is now doubling every 6 monthsnote3 as the Scaling Laws paradigm has taken over form Moore’s Law (2x in 18months) and the new currency is tokens per dollar per watt. There is currently no dominant marketplace player for AI agents.
Why Kondor and Ora?
In the fast-developing AI technology space, Kondor Directors identified an opportunity to create a unified marketplace for businesses and AI agents which will be called Sundae_Bar (S_B). AI agents have emerged as one of the transformative segments in the AI market growth and refer to a software system or program capable of autonomously performing tasks on behalf of a user or another system by designing its workflow and utilising available tools ("AI Agents"). Ora's infrastructure can support the scaling of Kondor's AI functions into a broader marketplace where developers create, share, and monetize other niche AI tools.
Following completion of the Acquisition, the Enlarged Group's objective is to create a dynamic platform that serves as a marketplace for AI Agents, enabling AI developers to market, manage, and sell their agents. This platform would provide essential tools for creating, customising, and monetizing AI agents, in order to offer a full-service solution for AI creators similar to how e-commerce platforms empower e-commerce entrepreneurs.
To enable a fast development of the sundae_bar platform, developers will have access to tools to scale and monetise AI Agents while giving businesses a trusted space to discover and implement AI solutions
The Kondor Directors believe that Kondor's iOS and Android application compatibility could rapidly be extended across a wider portfolio of AI Agents, ensuring the platform is accessible on mobile and desktop environments. Ora's infrastructure can support the scaling of Kondor's application functions into a broader marketplace where developers create, share, and monetise other niche AI tools.
Marketplace is fragmented
There are numerous avenues to finding AI tools by going to known vendors with the large players throwing huge advertising budgets at it. Equally developers have the ability to create but not necessarily have a way of finding the organisations or individuals who would match with their needs. Some models are totally free whilst others seek subscriptions. There is also a proprietary feel in that you are hooking up into distinct networks, with little homogeneity over systems, payment methods, hosting ability etc. Sundae_Bar is a unique combination of AI development ability, with open-source capabilities, on a proven platform. There is huge potential here to be a first mover with other players trying to dominate and attract to their own model platforms directly. The potential to be a Shopify or Betfair exchange, true disintermediation, is significant.
The question is where are you on that AI Agency Gapnote4?
Questions remain about data privacy, bias, and ethics but AI agents are not going away. The programming will only improve whether it is machine learning, automation or smarter algorithms. They will be closing that agency gap.
Wiz reportnote5/5a 85% of public cloud users are using some form of AI whether self-hosted or managed. 8 out of 10 of the most popular hosted technologies being associated with open-source models. This mix reflects the industry's dynamic nature, where open-source tools and proprietary models coexist and often complement each other in cloud environments. It also leans towards experimentation and development with a bias towards early adopters deploying fewer than 10 instances of AI rather than large scale production use.
This favours a centralised marketplace which if it can allay fears over security and bad actors, perhaps being at the forefront of creating some form of standard and governance for AI models then Sundae_Bar usage could surgenote5a. Sundae_Bar comes with scale of 60,000 AI enthusiasts already signed up with Kondor AI. Ora’s platform is enterprise-grade so will be attractive to the high-end player as well as newbies. Many on the board have Ai architecture building experience. Crucially, there are built in compliance systems like Know your client KYC and Anti-Money Laundering AML features.
Agent.Ai is a San Francisco platform focused on development but also trying to match users with developers. This is a free model which utilises credits for sales as a monetisation element. The designer of this is Dharmesh Shah... and he is also Chief Technical Officer (CTO) and co-founder of HubSpot which is an all-in-one sales and marketing CRM (customer relationship marketing) digital customer platform providing solutions to help organisations grow with their clients. Agent.ai is a natural progression for them and is Shah’s baby.
Chatspot.…they are the AI add on to Hubspot under the banner Breeze Copilot but Shah’s real project is agent.ai which has gone from 100k users in January to 500k by February which is 10X growth in 6 months. Also c.5k have tried to use the low-code tool for building an agentnote6.
This demonstrates the interest in this sector and strategy. Sundae_Bar will present as an immediate way to monetise for the users.
GOOGLE/Alphabet were looking at buying Hubspot which from afar seemed similar to when MSFT bought Github
Why did they pull out? They never reached terms to go to the due diligence stage. No reasons given but was around the time that 50 customers had their accounts hacked. They had 205k customers at the time in July 2024 but was possibly the final nail on an unsure strategy. A $25bn takeover would have been Google’s largest acquisition. Not to be deterred they they went back to Wiz, the cyber-security firm that specialises in improved cloud security and the ability to use multi-clouds, for $32bn all cash dealnote7.
Due to privacy laws changing Google needs to own first party data, with them phasing out 3rd party cookies by Q3 2024note8.
This is all included to show that data, upscaling your offering and making a full suite of tools available is the direction of travel with the big players. Collaborative platforms with enhanced security features is the trend.
Swarmzeronote9 is a Canadian private company and one of many players out there coming to the same conclusion that collaborative, democratised platforms are the way to go. They have gone down the subscription path. This is in the mix for S_B. Investor wise there isn’t really a listed version of what S_B is proposing for direct investment; you have to buy into a large player with multi-business lines.
Github was the original software/developer platform that saw monstrous growth. They started to solve a problem in the Linux programming world and recognising that their solution Git was difficult to use they set up GitHub to make it easier to share and collaborate on code using GIT. This was back in 2008 but very quickly morphed into the home for software developmentnote10/10ab seeing exponential growth into 2018 when MSFT bought it. Satya Nadella had taken over from Steve Ballmer in 2014 as CEO at MSFT and openly stated Microsoft’s love for Linux, or open-source communities. At the July 2024 earnings call Nadella said that the coding assistant Github Copliot accounted for 40% of GitHub’s revenue growth for the year and that is bigger than the whole of GitHub when they acquired it in 2018note11.
This shows how this concept can proliferate. Sundae_Bar will look to be the AI Agent expert version of this.
A digital but non-AI specific model example is Shopifynote12 which allows users, either first-timers or established businesses are helped to start, scale and succeed in the e-commerce world. They too give you an opportunity to build your offering with no coding experience required and an ability to monetise straight away with a secure payment system. You are able to create a dedicated space for your product and sell it on multiple platforms just like Sundae_bar’s mobile capabilities. Shopify allows 3rd party apps and AI integrations just as S_B’s will be language model (LLM) agnostic and allow the developer to access and utilise all AI models. Shopify has a checkout designed to convert browsers into buyers something which S_B will look to replicate for the AI world.
S_B will benefit the proliferation of off the shelf product matching alongside bringing the bespoke players together.
AI models have tightened up. Fewer hallucinations are being returned, when they appear to make things up. They are growing more capable and writing better. Effectively getting better faster.
Moore’s law is smashed from 18m to 6m to double power.
OpenAI GPT o1 can reason at PHd levelnote 14 and show you how it made the analysis.
Multi modal models are here using Text, Audio, images and video.
Software that uses AI and tools to accomplish a goal requiring multiple steps.
Uses AI models and tools and critically remembers its work.
Agents will be building blocks for other agents. This will allow many to become developers with little coding experience and see others help them develop their AI models. A drag and drop nature just like Workflows developed.
The future is hybrid teams. AI making humans stronger
Marketplaces are designed to encourage development and coding improvements to ultimately match traffic with qualified leads. That is worth paying for.
Agent.ai had 47k users and 1700 developersnote13(22mins) when he was showcasing at Inbound in Sep 2024. Kondor had 60k users and expects c 6k to be developers before any pre-launch marketing. Given the interest Shah’s baby has now seen, a ten-fold increase in users at agent.ai, then these numbers for Sundae_Bar could easily explode.
Agents are now everywhere across all sectors so agent ai is a huge subset of AI and many think the dominant one.
There is an agent for that… think of agents as the new apps but they need a marketplace. If you are fortune 1000 company you will be tied up with a large company subject to your wallet size.
The future is for everyone else that is looking to monetise and grow. 2023-25 development in AI was momentous but 2025-7 is likely to go bigger. A platform needs smart crm, smart engagement apps and an agent ecosystem.
You see it mentioned a lot on advertising and in company reports but what is the real driver; is it just hope after a lot of up-front spend? Some investors may remember the mobile spectrum overspend on 3-5G and still shudder.
Just how revolutionary is agent AI?
MSFT CEO Satya Nadella has made some huge claimsnote15 about the reversal of the Saas domination and perhaps ending the whole application software model. He sees a reduction in apps/software/SaaS and an increase in AI-powered experiences. An AI agent is a UI on the database. An AI tier. Just tell it what you want and if it can’t do it already it will write code to make it happen. AI doesn’t care what your backend looks like so build the most cost-efficient database and free the agents to do the work.
Copilot with an agent looking at the data. Is there even a need for one of Microsoft’s big apps, excel, which is a UI on top of a database? Just instruct the copilot agent the required end result and it will write the python code to execute it.
Potentially then, a complete restructuring of the SaaS industry. You don’t need the apps.
It’s just a database plus UI. Why pay for a UI when you can have the agent instead. If this is one of the biggest player’s views, and one with a lot invested in traditional apps, then his views potentially have weight. This has the potential to further increase the developer universe if suddenly some mainstream products can be replaced with agents or an agent overlay. Either way the penetration numbers for agents and an agent AI marketplace could actually be bigger than first thought.
Given the commonality of shareholders between the two companies Kondor had chosen Ora as an acquisition target. On 23 August 2024 Kondor and Ora, pursuant to Rule 2.4 of the City Code, made a joint announcement of a possible offer for Ora by Kondor. April 16th was the AGM vote on the takeover. The common shareholdings suggest that the deal will easily meet the more than 50% threshold and there is little rationale as an Ora shareholder to not vote for the deal since the Ora Board has already suggested that the company is not a going concern in its current carbon market. They had the right model but the risk of the Carbon credit market not proliferating effectively overtook events.
There is a potential for this to move positively since there is strategic rationale and cost synergies from the deal, with the potential to be a significant destination as an AI marketplace.
The AGM result saw 61.98% of acceptances and has been declared unconditional. Once 90% is achieved Kondor will enter into compulsory purchase of the remainder, under the Companies act. Anyone who doesn’t deal into that will be left as a minority holder and possibly unable to sell again anywhere near the Takeover offer terms. In all likelihood this should be fully participated in. The latest updatenote17 showed acceptances of the deal at 86.33%
"As at 31 July 2024, the Ora’s current liabilities exceed the current assets by £31,865. As part of the going concern assessment, the Directors have considered the implications of the potential transaction with Kondor AI Plc whereby Kondor AI Plc purchases the entire share capital of the Company and the ‘enlarged’ entity lists onto one of the London Markets. In parallel with this transaction, it is anticipated that there will be a further fundraise by Kondor AI Plc of an amount sufficient to enable the ‘enlarged’ entity to continue in operational existence for the period to at least 31 January 2026."
So here we are with the announcement to list on Aim as Sundae_Barnote18 and an IPO raise of £2m at 8p per share which will see 25m new shares added. For the most conservative read of the shares in issue we are assuming a 100% conversion of Ora shares which would be 386,730,050 Kondor AI shares + 25m issued for the IPO listing of Sundae_Bar for 411,730,050 shares outstanding
Model based on $30 transaction costs that are repeatable but I’ve assumed monthly by estimated developers in the first instance, payable by the developers at a 2% transaction fee rate to Sundae_Bar
When growth comes through it should have the exponential profile of an adoption curve model, the hockey stick curvenote16
Github growth and agent/ai growth have both seen exponential user growth at the start of their collaborative platform journeys. We could see multiple developers having high factor hit rates but to be conservative the assumption is the exponential growth of users, with a readthrough percentage of developers. Eventually, we will be able to model the full suite of subscriptions, transaction fees and Placement fees the company anticipates.
The pathway from free to paid is tied to demonstrating the value up the chain. Of being plugged in for the important iterations of development. The value for S_B could be huge if able to convince the sector that they are the marketplace where the whole sector wants to play not just a particular subset. There is bound to be a goto bias towards the large advertising spenders but the empowering nature of the agent world for the users should start to focus on availability of developer, platform and technology flexibility, along with the potential to spot trends centrally and not in a tied-up nature. Deepseek updates and subsequent downloads show how important it is to be ready to adapt.
A game changer would be if the platform gains acceptance, it sees the large players place themselves on it as eventually the scale of the market will demand a portion of revenues generated from a centralised source alongside directly sourced client flow. AI agents by definition will perhaps recognise the need to source the marketplace themselves, particularly if it demands a governance and security premium amongst the users.
The success of platforms like GitHub was driven by the easy graphical interface that beginners and professionals alike could interact with. This made exchanging code, reviewing each other’s project and managing software code in real-time an easy process. It is basically free to use to attract traffic and they monetise by offering hosted private repositories (unique locations for project files), premium subs packages to teams and enterprises, and by adding fees when users buy third-party apps on their platformnote10b. This latter part is the first charging structure S_B is focusing on with a 2% transaction fee. Prioritise developers and attract traffic.
As partnerships with developers grow the prospect of commissions for extra services develops. Ultimately, as any strong marketplace discovers, there is power in the search data and advertising fees for platform placements.
Subscriptions are a popular model in the space and that will have to be worked out in context of the benefits of the democratised platform. The inexperienced user or small business user particularly may be attracted here for extra tools and experience. The main thrust though is monetising the developer and making a commission. Should Sundae_Bar become a marketplace, where trust in it is firmly established, then traffic to the site could accelerate much quicker than originally thought.
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 £/$ 1.3305 as of April 23rd, 2025.
Now if we assume that the agent.ai trend is replicable and the 10x increase in users possible in Q3/4 of 2025 we could imagine 500k Sundae_Bar users by year end. Without changing any of the other earning parameters that has a material impact on the DCF below:
We are pre-launch and pre-revenue, in a new idea, albeit in a space that is attracting users and the concept is in demand from the little man up to the Mag7. Although it would be ideal to fly out of the blocks the success of this model will rely on trust being built and security concerns alleviated. Sunadae_Bar is coming from a strong position platform wise and the payment infrastructure should be defining, not to mention mobile roll-out. A successful venture could grow at warp speed in what is looking like a potential disruptor moment for the software industry. They are placed to be where the future is, agentic AI, and trying to help lead how the industry develops from a standardisation angle.
As a newcomer or someone not experienced in coding the platform should allow you to come up to speed, match with your developer or learn enough to do it yourself, alongside your agent AI teammate of course. Lots of fretting around in the world at large about lost jobs but the idea is to create more valuable humans who can analyse more rather than manual tasks.
A one-year price target between the two presented Fully diluted DCFs of £0.31 - £1.13 seems reasonable and I will suggest 60p in the hope that the current market tariff turmoil is sorted and we can return to looking at the global growth phenomenon that agent AI is. The $ exchange rate may even stop falling against the pound and boost any translated earnings.
Some risks to think about:
The market is moving too quickly and the risk of agents replacing developers leads to large scale mistakes reducing confidence in the idea. Unlikely to be the case for too long. Agentic – SoftWare Engineers (A-SWEs) are on the way but will still need monitoring.
Lack of homogenised standards and security issues. Many fear bad actors putting in code or corrupting databases in order to force AI agents to react maliciously. Indeed, we saw mass adoption of DeepSeek’s model recently but Wiz, who was recently bought by Google, identified some serious security glitches along with some Nvidia chip issuesnote5a.
IF Sundae_Bar can establish itself as truly agnostic and not tied to any large players then logic would suggest it stands out as the platform of choice. Agents may well recognise this too and have it on the search list for useful agents too.
Now that sounds like sprinkles with raspberry sauce to me.
DOWNLOAD FREE REPORT
Fill out the short form below and get this free report sent direct to your inbox.
Note 1 STATISTA.COM AI market size worldwide from 2020-2030 Date 28th Nov 2024
Note 1a LITSLINK.com / Blog / AI agent statistics Date 6th March 2025 Sergey Antonyuk
Note 2 PRNEWSWIRE.COM Ai agent market worth $47.1bn by 2030 Date 12th Sep 2024
Note 3 Publish.obsidian.md Computing power is now doubling every 6 months as the Scaling Laws paradigm has taken over form Moore’s Law and the new currency is tokens per dollar per watt Date Oct 21. 2024
.
Note 4 TheJournal.com / articles / mind the agency gap Date 23rd October 2024
Note 5 CampusTechnology.com / Articles / Report 85 of Organisations are using some form of AI Date 28th Feb, 2025
Note 5a Wiz.io / reports / The state of ai in the cloud 2025 (downloadable)
Date April 2025
https://www.wiz.io/reports/the-state-of-ai-in-the-cloud-2025
Note 6 CXTODAY.com Hubspots prospective ai agent platform races past 500k users Date Feb 3rd 2025
https://www.cxtoday.com/crm/hubspots-prospective-ai-agent-platform-races-past-500000-users/
Note 7 Blog.google.com Google agreement acquire Wiz Date March 18th 2025
https://blog.google/inside-google/company-announcements/google-agreement-acquire-wiz/
Note 8 LinkedIn.com Why Google Would Acquire Hubspot. Brendan King Date April 9th, 2024
https://www.linkedin.com/pulse/why-google-would-acquire-hubspot-brendan-king-ajwzc/
Note 9 Swarmzero.ai Company webpage
Note 10 Blog.Getlatka.com Github revenue Date Mar 2nd, 2025
https://blog.getlatka.com/github-revenue/
Note 10a Blog.gitbutler.com Why Github actually won
Date September 2024 by Scott Chacon
https://blog.gitbutler.com/why-github-actually-won/
Note 10b Productify.substack.com How Github grows and makes money
Date July 6th, 2024 by Bandon Jot Singh
https://productify.substack.com/p/how-github-grows-and-makes-money
Note 11 ITbrew.com Microsoft earnings show role of Copilot tool in revenue growth Date Aug 13th 2024
Note 12 Shopify.com / blog / What is shopify? Date Mar 5th, 2025
https://www.shopify.com/blog/what-is-shopify
Note 13 youtube.com / The Future Of AI Agents With Dharmesh Shah | INBOUND 2024 Date October 2024
https://www.youtube.com/watch?v=IityUpVVD38
Note 14 Venturebeat.com / Forget GPT-5! OpenAI launches new AI model family o1 claiming PhD-level performance Date Sep 12th, 2024
Note 15 Youtube.com Microsoft CEO’s Shocking Prediction: “Agents Will Replace ALL Software" Matthew Berman Date Jan 2025
https://www.youtube.com/watch?v=uGOLYz2pgr8
Note 16 Techtarget.com Hockey Stick Growth Date October 2022
https://www.techtarget.com/searchcustomerexperience/definition/hockey-stick-growth
Note 17 LondonStockExchange.com Kondor AI Plc – Acceptance Level Update Date 25th April 2025
https://www.londonstockexchange.com/news-article/market-news/acceptance-level-update/17005173
Note 18 Announcement re Sundae_Bar IPO listing on Aim Date 12th May 2025
https://www.londonstockexchange.com/news-article/market-news/schedule-1-sundae-bar-plc/17030817
Nothing in the above article should be considered investment advice.
Small Cap or Aquis listed companies can be highly illiquid making them difficult to sell at the quoted price, and in some cases, it may be difficult to sell them at any price. Small Cap or Aquis listed companies can have a large bid / offer spread which means there could be a large difference between the buying and selling price. Companies listed on the Aquis market can be highly volatile and are considered high risk speculative investments. The value of your investment can go down as well as up, your Capital is at risk you may not get back the amount invested. Investments in IPO’s & RTO’s involve a high degree of risk and are not suitable for all investors. All investments made into an IPO, RTO in a secondary issue should always be made solely based on the information provided in the relevant prospectus and any other supplementary documentation. Please ensure that you fully understand the risks involved. If in any doubt, please seek independent financial advice. This document is published by Clear Capital Markets and does not constitute a solicitation or personal recommendation for the purchase or sale of investment. The investments referred to may not be suitable for all investors. Any data or views given should not be construed as investment advice. Every effort is made to ensure the accuracy of the information, but no assurance or warranties are given. Clear Capital Markets Limited is authorised and regulated by the Financial Conduct Authority FRN 706689.
Clear Capital Markets Corporate Broking acts as a Corporate Broker to sundae bar Plc and holds warrants and shares in the company. Employees and/or directors of Clear Capital Markets may deal in shares of sundae bar Plc for their own personal accounts. These scenarios may give rise to a conflict of interest where Clear Capital Markets also provides clients with an advisory service for transactions involving sundae bar Plc. The firm has established Conflicts of Interest (“COI”) and Personal Account Dealing (“PAD”) policies to mitigate the risk of a conflict causing damage to the interests of its clients. The measures taken include (i) enforcing minimum holding or ‘lock-in’ periods; and (ii) requiring internal review and approval from the compliance department for employees or directors entering into personal transactions involving sundae bar Plc. The COI and PAD policies are available upon request. Before Clear Capital Markets proceeds with a placing, a number of factors are considered including: liquidity of stock, company diversification, market capitalisation and potential news flow. Only once minimum criteria are satisfied would we elect to proceed. Any remuneration payable to Clear Capital Markets has no bearing on whether it proceeds with a placing. These administrative controls mitigate the risk of a conflict causing damage to the interest of a client, but the inherent risks of this business model cannot be eliminated. Accordingly, Clear Capital Markets is required to disclose this conflict to help clients to assess the service that they are being offered in light of Clear Capital’s own interests, and to decide on the extent (if at all) to which they will rely on, or proceed with, the service.
We will reach out for a quick introductory fact find to see how we can best serve your specific needs and arrange a convenient time for a consultation with a member of the appropriate specialist team.
An Advisory Broker will contact you for a deeper fact find and share information on how we can best enhance your portfolio.
We will conduct a full RCP in accordance with FCA regulations to ensure that your appropriate risk profile is set. You will then have access to an extensive suite of investments, supported by your dedicated Advisory Broker.
