The DayOne network — Code of Conduct

The DayOne network is a curated, AI-forward network of South African developers, creatives, and specialists. It works because of who's in it and how we treat each other and the clients we serve. This document is the standard we hold ourselves to. It's short on purpose. If you're the kind of person we want in the network, most of it will already be obvious.

DayOne exists to create opportunities for the network. Those opportunities only keep coming if the work is good. The quality of the AI services we deliver is the whole game — it's what turns one client into three, and one project into a reputation the network can trade on. Every member's output is, in a real sense, the network's pipeline. Treat it that way.

It applies to everyone — founding members, members, and anyone acting on behalf of the DayOne network.

• Quality. The work carries the network's name. Ship things you'd put your own name on.
• Collaboration. Help peers when you can. A network of people who actually back each other beats a directory of strangers.
• Fairness. Credit people honestly. Compete on merit. Don't cut someone out of work they helped win.
• Accessibility. We're here because good tech and good clients used to be a privilege of the well-connected. Don't pull the ladder up behind you.
• Partnership. Treat clients, peers, and partners as people you'll work with again — because you probably will.

Engage in good faith. Disagree on substance, not on people. No harassment, discrimination, or abuse — on the basis of race, gender, sexuality, religion, disability, language, or anything else. This is non-negotiable and not subject to "it was a joke."

What's shared inside the network — work in progress, client situations, someone's half-finished idea — stays inside the network unless the person sharing it says otherwise.

Do what you say you'll do, or tell people early when you can't. Missed deadlines happen; silence is what damages trust.

Hold a quality bar regardless of how the work was made. The tooling is your business; the output is the network's reputation.

If you bring a peer onto a piece of work, agree on terms and credit up front. Don't renegotiate someone's share after the fact.

The DayOne network is AI-forward. That means we expect members to use AI as a tool and an accelerator — and to use it well.

• Quality is the bar, not effort. AI-assisted is fine. Slop is not. If you'd be embarrassed for a peer to see how little you checked it, don't ship it.
• Be honest with clients where it matters. Don't misrepresent what's AI-generated when a client has a reasonable need to know — for example where accuracy, rights, or liability are at stake.
• Respect rights and data. Don't feed confidential client material into tools in ways the client hasn't agreed to. Don't pass off others' work, AI-laundered or not, as your own.
• Know the limits. Part of being AI-forward is knowing where AI helps a client and where it doesn't. Advise accordingly. Don't sell AI for its own sake.

Our value to a client isn't "we use AI." It's that we know where AI belongs in their business and where it doesn't. That judgement is what they're actually paying for, and it's the standard the network holds.

• Lead with the business, not the tool. Start from what the client is trying to achieve, where AI can enable their operations, and the strategy behind it. The tech serves the goal, not the other way around.
• AI is exceptional at early-stage and prototyping. Used well, it gets a client from idea to working prototype faster and cheaper than anything before it. Lean into that strength.
• Say so when AI isn't the right fit. When AI won't sufficiently meet the client's needs, we tell them plainly — even when it costs us the easy sell. An honest "this isn't where AI helps you" earns more trust, and more future work, than a forced one. Misselling AI damages the client, the member, and the network.

Client confidentiality is absolute unless the client agrees otherwise. Showcase your work only with permission, and anonymise where needed.

Credit collaborators by name. If a peer contributed, say so. Take credit for your part, not theirs.

The DayOne network is not a job board and not a recruitment platform. Don't treat it as one.

That means: no CV spam, no scraping the member list for leads, no using the network primarily to harvest contacts for something happening elsewhere. Connecting a peer to good work is welcome and encouraged. Mining the network is not. If your main reason for being here is lead extraction, you'll be the wrong fit, and it'll show.

When you work with a UK or EU client through the network, you're representing more than yourself. Professionalism, clear communication, and delivery are how SA talent earns the next ten roles for the people behind you. Treat every client like a reference you'll need later.

Founding members carry an extra responsibility: you help build and curate the DayOne network, and you set the tone. The standard the network holds is the standard founding members model. Vet thoughtfully. Welcome newcomers. Flag what's not working. You asked for a say in what this becomes — this is the work of having one.

Partnerships with other networks, organisations, or platforms go through DayOne directly. Don't commit the DayOne network to a partnership, or represent it as endorsing one, without that conversation having happened.

Membership is vetted, and it can be reviewed.

• First, a conversation. Most issues are misunderstandings and get resolved by talking.
• Then, a warning. Where a real line was crossed.
• Then, removal. For serious or repeated breaches — harassment, dishonesty, breaking client confidentiality, or treating the network as a CV farm after being asked to stop.

Some things skip straight to removal: harassment, fraud, or anything that puts a client or a member at genuine risk. We'd rather lose a member than the trust of the network.

If you think a decision was wrong, you can raise it with DayOne directly.

The DayOne network is early. This document will get sharper as we learn what the network needs. Founding members help write the next version. If something here is unclear, missing, or wrong — tell us. That's the job.