How to integrate AI into your business: use cases, risks and automation
How to integrate AI into your business: start from a specific task, know when a rule is enough, choose between an automation tool and custom software, and know the risks.
Updated on 16 min read
Sébastien Nobour
Founder of DEVEDANOS
Have you raised funds or secured start-up funding, or are you in the middle of doing so? Does your product not exist yet, or only as a mock-up, and do you need someone to build it and to make the technical decisions that will shape your company for years? You may be thinking of hiring a chief technology officer, or CTO, on a permanent contract.
An employed CTO builds and decides. They are paid a salary rather than in equity. You pay that salary every month, whether or not your product finds its customers. You also have to find them, then judge their skills without being technical yourself. In our view, the best time to make this hire is when your first users are already using your product.
CTO, or chief technology officer, is a role: whoever holds it leads the technical side of your product. The title says nothing about how they are tied to your company, nor how long they have held it. At the very beginning, writes Fred Wilson, an investor at the American fund Union Square Ventures, a startup has neither a CTO nor a VP of engineering yet: its technical co-founder plays both roles.
The technical co-founder takes equity and shares your risk. They often hold the CTO title, and finding a technical co-founder means giving up part of your company.
The employed CTO is tied to your company by an employment contract. They work for you alone, for a salary paid every month: they are the subject of the rest of this article.
The fractional CTO works for several companies. They make the technical decisions, and someone else builds the product: this is one of the four forms a technical partner takes.
The founding engineer is one of the very first developers hired. They are an employee, often granted a small share of the company’s equity, not a co-founder.
The product engineer is a developer who takes ownership of the product as much as of the code: they talk to users, and take part in deciding what gets built first.
A freelance developer or an agency can also build your product, under a commercial contract. Before you sign, check what that contract must cover.
At first, an employed CTO builds the product themselves. They write the code, choose the technologies, decide what gets done first and what will wait. As the company grows, they hire the technical team and lead it.
They work for you alone, full time. Software they develop in the course of their duties belongs to your company (section 11(2) of the Copyright, Designs and Patents Act 1988). For a co-founder or a service provider, you need a written assignment of rights, the document by which the author of code transfers their rights in that code. For an employee, the law is enough.
They do not decide alone what your product should do. Marty Cagan, a product management specialist and author of the book Inspired, identifies four risks in building a product, and entrusts only one of them to engineers: feasibility. Whether customers want the product, and whether it fits the rest of the business, is the job of whoever is responsible for the product, and in a young company, that is often you.
So a CTO cannot know on your behalf whether customers want your product. The research firm CB Insights analysed 111 startups that had shut down, in a 2021 study reported by the SmartCompany website. No market need was among the reasons cited by 35% of them. Before you hire, put the question to your future customers.
The salary. The salary of a startup CTO depends on their experience and on your sector, and there is no standard rate for the role. Set the amount you can carry before you make an offer, and ask recruitment agencies and comparable companies in your country what they pay for this profile.
Employer contributions. They come on top of the gross salary. In the 2026 to 2027 tax year, the employer pays National Insurance at 15% on pay above £5,000 a year, and at least 3% into the employee’s workplace pension, on earnings between £6,240 and £50,270.
Recruitment. If you go through a recruitment agency, its fee comes on top. It is negotiated with the agency, so ask for it in writing before you instruct it.
The annual cost to the employer is therefore the gross salary plus National Insurance and the minimum pension contribution, and agency fees, paid once, come on top. We compare this cost with that of a developer.
Divide the funds you have available by this monthly cost: you get the number of months you can pay your CTO before your product brings in money.
In an article published by Journal du Net in June 2026, non-technical founders explain how they chose their CTO. Amélie Schieber, president of Tiime, a French fintech that offers an accounting platform to freelancers, entrusted the search for her CTO to a recruitment firm. She had the technical skills of the candidate the firm put forward assessed by the firm itself, and by her company’s technical teams. The article also points out that a profile suited to launching a product does not always suit its growth: tell the candidate how far you want to take your product, and ask what they have already scaled to that size.
If you have nobody to assess a candidate, get help: from a specialist recruitment firm, from a developer you trust, or from a fractional CTO. And ask the candidate the question you would ask a co-founder: what software have they put into production, then kept developing over several years?
How long it takes to part ways. UK law has no statutory probation period; employers set their own, of any length (Acas). The notice period is whatever your CTO’s contract says, and the law only sets a floor: at least one week’s notice from one month of service, then one week for each year of service between two and twelve years, and twelve weeks from twelve years. Your CTO’s contract can provide for longer. Under the Employment Rights Act 2025, protection from unfair dismissal is due to start after six months’ service from 1 January 2027.
The cost of a failed hire. A French article on HelloWorkplace works through a case: an employee who costs the employer €50,000 a year, contributions included, leaves the company after six months. Adding up the salary and contributions already paid, onboarding and training, lost productivity and recruiting a replacement, the article puts this failed hire at €43,000. This example concerns a different profile, and a CTO costs the employer more: this figure gives only an order of magnitude. A CTO’s successor will also have to take over the code they wrote. Keep enough in reserve to absorb a failed hire.
Without a CTO, you can also hire a developer first.
A developer writes the code, tests it, puts it into production and fixes it. In a startup without a CTO, they also make decisions a CTO would make elsewhere: the language, the hosting, how the code is organised, what they check automatically and what they leave aside. These decisions stay with your company for years.
Without a CTO, there is no technical team yet to give its opinion on a first developer’s work. You can still judge it in three ways.
What your users do with it. They use the software, or they do not. They report what is missing and what does not work. That judgement is yours, and it takes no technical skill.
How often, and how reliably, changes go into production. DORA (DevOps Research and Assessment), Google Cloud’s research programme on how software is developed and put into production, tracks five metrics. You can ask your developer for three of them every month, without reading a line of code: how many times a change was put into production, how many of those releases broke something and needed immediate intervention, and how long it took to restore service. DORA’s research regularly shows that “speed and stability are not tradeoffs”: a team that releases often does not pay for it in outages.
Dave Farley, one of the pioneers of continuous delivery and co-author of the book Continuous Delivery, asks a question in Modern Software Engineering that you can borrow: “How long for a single-line change to go from ‘idea’ to ‘working software’?” Put it to your developer, and ask them to show you the path that change follows.
An outside technical eye. Code quality, security and architectural choices cannot be seen from outside the software. Judging them takes another professional: a fractional CTO, or another developer who reviews part of the code at regular intervals.
That outside eye matters all the more because your developer will probably write part of the code with AI. In the 2025 Stack Overflow survey, run by the developers’ question-and-answer site, 84% of respondents use or plan to use AI tools to develop. In the same survey, 46% of respondents distrust the accuracy of what these tools produce, against 33% who trust it. For Dave Farley, writing code now costs little, and it is verification that takes the time. Ask your developer how each change is checked before it goes into production.
In our view, what protects a first developer’s work when there is no CTO is continuous delivery. From their first month, before your customers depend on their software, check that they know and apply its principles: a deployment pipeline as Dave Farley teaches it, which reruns automated tests on every change, small changes delivered continuously, and a software design that mirrors how your business works.
When a first developer is the right answer. Your product is already in use, and it has to change every week at your customers’ pace. That is work for one person full time, for years, and an employee who knows your business from the inside can commit to it for the long term. Their work will earn your company more than it costs, and you can afford their salary. And someone, a fractional CTO or an outside developer, can check that they apply the principles of continuous delivery. In that case, hiring a first developer is a good choice.
Paul Graham, co-founder of the startup accelerator Y Combinator, wrote in October 2015: “Hiring too fast is by far the biggest killer of startups that raise money.” He adds that founders who have just raised money are often encouraged to hire more than they need by the investors who funded them.
In France, according to two players in startup financing, investors are asking for more evidence before investing. Grenoble Angels, a network of individual investors in the Isère department and a member of the France Angels federation, observed in the second half of 2024 “higher expectations in terms of market traction, and therefore revenue”. In July 2025, Nicolas Forey, chairman of the consultancy In Extenso Innovation Croissance, commented in Le Journal des Entreprises on the fundraising barometer of the accounting network In Extenso, which compares the first half of 2025 with 2023. According to him, young startups find it harder to complete their first fundraising round, the seed round, which was still holding up in 2024. The newspaper adds that investors, now more cautious, “prefer to refocus their activity on more mature companies”.
An idea alone can still raise money from some investors. Y Combinator states in its FAQ that, on average, 40% of the companies it funds in each batch are still just an idea. What you do with that money remains your decision.
In our view, the money raised should go first towards checking that customers want your product. Hiring an employed CTO before that check means paying a senior salary every month for a product nobody yet knows will find its customers. Once your first users are using it, you hire for needs you have observed, and you can explain them to your future CTO.
What evidence do you have today that customers want your product? If the answer is still an intention, check it first with an MVP, the minimum viable product. It is not a first version of the product, but the smallest test that checks whether the idea is worth building.
Your employed CTO’s code belongs to your company, but what they know about the software does not. As long as they are on their own, all knowledge of the software rests with them. If they leave the company, it leaves with them, unless it has been written down and handed over. Their notice period, set by their contract, is the time you have for them to hand it over. Ask them from the start to write down how the software works, as they go.
Also open the software’s accounts in the company’s name, not your CTO’s: the domain name, the hosting provider, and the repository where its source code is kept.
Only a takeover trial, carried out once before you need it, proves that another developer can take over the software: they receive the code and the documentation, install the software, then put a small change into production. If they succeed, the knowledge is in the company.
A CTO brings their skills, and also a way of deciding what gets developed, and in what order. With a predictive approach, the list of features is fixed at the outset, with a schedule and dates. Yet that list is written before the software exists, on the day you know least about how your application’s users will use it. With an adaptive approach, the software gets into their hands early, and how they use it defines the next features.
During recruitment, ask the candidate how they will decide what gets developed first, and what they will do on the day your users ask for something other than what was planned.
An employed CTO is justified when four conditions are met. Users are already using your product, or customers have committed to it. Your funds can cover their salary for as long as it takes your product to bring in money. Their work will then earn your company more than it costs. And technology is at the heart of your product: technical decisions are made every day, and call for someone full time. They are also the one who will hire and lead the technical team as it grows.
For Y Combinator, the founding team should have the skills to build its product itself from the start. If your product is first and foremost a technology, software whose value lies in its technical difficulty, this advice applies: hiring a CTO early, or taking them on as a co-founder, is then the right answer.
If you cannot pay a salary, the question becomes one of a technical co-founder. If what you mainly need is decisions and an eye on your service providers’ work, a fractional CTO may be enough.
Your product may also first meet a need you know through your trade or your customers. The technology then serves that need. In our view, you can have your product built now, with code that belongs to your company, and put it in the hands of your first users. During this period, the technical skills sit outside the company. Y Combinator advises the opposite in its FAQ: that the founding team should have the skills to build its product itself, rather than hand it to a contractor. For a product whose need comes from your trade, we think differently: its difficulty lies first in that need, which you know, and a contractor can build it as long as the code belongs to your company. You then hire your CTO on the strength of a product already in use, and they take over software whose code belongs to you.
A CTO hired once your users are already using the product knows what needs to change, and you know why you are hiring them.
Until that hire, we make the technical decisions your product calls for, as a CTO would: the technologies, the architecture, how it is tested and put into production. We align them with your objectives, and the decisions about your product remain yours.
At DEVEDANOS, we develop your custom software, usable within a month, for a fixed monthly budget, and we adjust it continuously as your business evolves. The budget is built with you, from your figures and your goals: together, we set the amount and the problem your application has to solve. With this fixed budget, you pay neither recruitment firm fees nor employer contributions. Your application’s users work with it from the first month of development, and how they use it determines the next feature or features. You can change direction whenever you want, with no contract amendment and at no extra cost.
Our approach is adaptive: it predicts what is predictable, the monthly budget and the development days per month, and does not try to predict what your users will do with a feature before they have it in their hands.
And so that what worked last month still works this month, we apply a zero-regression guarantee: every regression is fixed and covered by an automated test so that it does not come back, at no extra cost.
The source code, infrastructure, deployment pipeline and documentation are your intellectual property once the first monthly invoice is paid. A trial period of up to 15 working days in the first month: no invoice if you do not want the deliverables.
You can terminate the contract in writing at any time, without justification. During the reversibility month, we support your team until they can build, test and deploy without us. On the day you hire your CTO, they find software already in use, with its code, its tests and its documentation. The pipeline goes on protecting every future change, whichever developer makes it.
Among other things, we build marketplaces, business applications and mobile apps.
Not sure whether to hire a CTO, or to have your product built first? Shall we talk?
How to integrate AI into your business: start from a specific task, know when a rule is enough, choose between an automation tool and custom software, and know the risks.
An Excel file shared between several users, errors, conflicting versions: when the spreadsheet has become an application, and what moving to a business web application changes.
Web application takeover: what to recover before changing software supplier, why tests come before any change, and when a rebuild is needed.