Skip to content

The Work You Don't See Behind an AI Roleplay Platform ​

Two people, countless responsibilities, and more than two years of learning.

By Edith | AICHIKI Community Manager

When you open an AI roleplay app, everything looks relatively straightforward. You choose a character, type a message, and receive a response. Perhaps you create your own characters, customise your conversations, or explore stories written by other users.

What you don't see is everything happening behind that experience.

My husband, Rudi, and I are the two people primarily responsible for building and running AICHIKI, an independent AI roleplay platform.

Rudi handles development, infrastructure, and the business itself. I manage our community, communication, creative content, and much of what happens between the platform and its users.

We don't have investors, a marketing department, or a large development team. We do, however, have a small group of moderators and beta testers who help us, and a community whose feedback continues to influence our decisions.

After more than two years of development, we're still learning what it takes to build and maintain a platform like this.

And I thought it was time to share some of the work that usually goes unseen.

It's More Than Writing Code ​

When Rudi first started developing AICHIKI, he thought the technical side would be relatively simple.

The basic concept of an AI chat application sounds straightforward: send a user's message to an AI model through an API, receive a response, and display it on the screen.

But making an AI model respond is not the same as making it participate in a coherent, ongoing story.

Characters need to maintain their personalities. Conversations need context. Memories must be managed, user sessions maintained, and the system needs to handle interactions far more complicated than a simple exchange of messages.

These challenges become particularly important in AI roleplay, where users may spend hundreds of messages developing relationships, creating fictional worlds, and following complex storylines.

When I asked Rudi what people might not realise about developing AICHIKI, he told me:

"You can't figure it out in a week of vibe coding. I've spent more than two years on AICHIKI and it's still not perfect."

I think that sums up the difference between how simple something can appear from the outside and how complicated it becomes once you start building it.

And writing code is only one part of his responsibilities.

AICHIKI operates through a registered Romanian company, which means taxes, legal requirements, business administration, and financial decisions also fall on his shoulders.

He handles infrastructure, deployments, app store requirements, and decisions about which ideas are technically, financially, and legally feasible.

He works with different AI models, learning their capabilities and limitations, and uses AI-assisted development tools such as Claude and Codex.

We even host and manage our own application infrastructure at home.

When you're employed as a developer, you can often finish a task, commit the code, and move on.

But when you're building your own product, the responsibility doesn't end when a feature starts working.

You have to consider whether people will use it, whether they understand it, how it affects other systems, and what happens when something goes wrong.

Rudi described another difference that stayed with me: when you own the product, it's always on your mind.

There's always something to improve, something to investigate, or another idea worth considering.

And I can certainly relate to that feeling, even though my work looks rather different from his.

My Side of AICHIKI: The People Behind the Messages ​

My responsibilities are centred around our users and the community we've built.

Most mornings, I start by checking Discord.

I read messages, answer questions, look into reported problems, and see whether anything requires Rudi's attention.

From there, I usually move between Reddit, email, social media, and AICHIKI itself.

Some days I'm preparing announcements, writing blog articles, creating promotional content, or organising community events.

Other days, I'm working on characters, updating existing content, selecting characters to feature, or helping users understand how certain features work.

And while I'm doing all of this, Discord remains open.

One of the parts I enjoy most is character creation.

I've spent a great deal of time developing characters, experimenting with their personalities, backgrounds, and memories, and observing how different AI models interpret the information they're given.

A good roleplay character needs more than an attractive portrait and a few personality traits.

Their history, behaviour, relationships, speech patterns, and reactions all contribute to how consistently they behave during a conversation.

Sometimes a seemingly insignificant detail makes a remarkable difference.

That experience has also helped me understand the difficulties other creators encounter, which makes it easier to offer practical advice rather than simply telling them what they should do.

But working with a community isn't only about creativity.

There are disagreements, misunderstandings, support requests, and occasionally situations that require difficult decisions.

Moderation is a responsibility I take seriously.

We have a small team of moderators who help maintain our Discord community, respond to everyday issues, and bring more serious matters to my attention.

Automated moderation also helps us identify potential problems, but I believe that context and human judgement remain important when making decisions that affect other people.

I handle decisions involving user bans, while Rudi becomes involved when an issue requires technical investigation or changes to the application.

What I've learned is that community management requires more than knowing the rules.

You need patience, consistency, and a willingness to listen, even when someone is frustrated or disagrees with a decision.

And sometimes the criticism you receive is exactly what you need to hear.

The Feature Rudi Didn't Want to Build ​

One of the things I appreciate about the way we work is that community feedback genuinely influences development.

I regularly bring user suggestions, complaints, and requests to Rudi.

Of course, that doesn't mean every suggestion becomes a feature. Some ideas aren't technically feasible, while others would be too expensive or wouldn't fit the direction of AICHIKI.

But user feedback plays a considerable role in deciding what we work on.

One of my favourite examples is the Continue button.

Rudi originally didn't want to include it.

To him, roleplay was supposed to be interactive. You write a message, the AI responds, and the story develops through that exchange.

He didn't particularly understand why someone would want to press a button instead of writing a response.

But users kept asking for it.

And eventually, he reconsidered.

What appears to be a simple button actually requires the system to instruct the AI to carry the scene forward without new dialogue from the user.

It isn't quite as straightforward as it looks.

Today, the Continue button is part of AICHIKI, and Rudi is happy that we added it.

I like this example because it demonstrates something important about building a product.

What makes sense to a developer doesn't always reflect what users need.

Sometimes the people using your application see possibilities or encounter frustrations that you wouldn't have considered yourself.

And if you're willing to listen, those experiences can make the product better.

A Small Team Doesn't Mean a Small Workload ​

Something people may not realise is how many different responsibilities need attention every day.

There's also an important detail about how we run AICHIKI: Rudi has a full-time job outside the platform.

AICHIKI is something he develops during his free time throughout the working week. After finishing his regular job, he spends whatever time he has available working on features, investigating problems, or managing the technical side of the application.

On weekends, AICHIKI essentially becomes his full-time job.

That's when he can dedicate longer stretches of uninterrupted time to development, updates, and the larger tasks that are difficult to fit into weekday evenings.

Meanwhile, my work follows a different routine.

I might be answering support messages, preparing a community event, writing an article, reviewing content, or discussing a moderation decision. Community management doesn't always follow a fixed schedule, especially when users can reach out at any time.

And we still have our family life.

We have two daughters, and our youngest still needs a great deal of my attention throughout the day.

Balancing community management with motherhood isn't always straightforward.

There are moments when a user needs help, an announcement needs to be prepared, or something unexpected happens, but I also have a little girl who needs her mother.

That's part of why having a reliable moderation team matters so much to me.

Our moderators help answer questions, manage everyday issues, and keep the community running when I can't be immediately available.

Our beta testers contribute by identifying problems, testing changes, and sharing observations that might otherwise be missed.

We don't always agree on everything, and that's perfectly normal. What matters is being able to discuss problems and work towards solutions.

I'm grateful for the time and effort they contribute.

Although Rudi and I carry most of the responsibility for AICHIKI, we certainly don't do everything without help.

One of Our Biggest Challenges Isn't Development ​

When I asked Rudi about the most difficult part of developing and maintaining AICHIKI, his answer surprised me a little.

It wasn't the programming.

It was finding users.

We have no investors and no dedicated marketing budget, so most of our efforts to reach new people rely on organic discovery, community participation, social media, and recommendations.

And there's another difficulty: reaching the right audience.

AICHIKI is built around character creation, storytelling, and sustained roleplay.

Not everyone interested in AI chat is looking for that kind of experience.

Some people want a quick conversation, while others are looking for a platform where they can develop characters, build relationships, and explore stories over time.

Our challenge is helping those people discover that AICHIKI exists.

That's one reason I've invested time in our social media presence, blog articles, community events, and educational content.

I want people to understand not only what AICHIKI offers, but also how they can use it to create better roleplay experiences.

Rudi occasionally joins the promotional efforts himself, including sharing AICHIKI on Reddit.

But marketing is still something we're learning.

You can spend months developing a feature, improving a system, or solving a technical problem, and still struggle to reach the people who might appreciate that work.

Building a product and getting people to discover it are two entirely different challenges.

And both require attention.

We're Still Waiting for the Millions ​

I couldn't resist including one final remark from Rudi.

When talking about the business side of AICHIKI, he admitted that we're not particularly profitable yet.

Then he added that we're still waiting for the millions.

I laughed when I read that.

But behind the joke is something very real about independent development.

We're investing time, effort, and resources into something we believe is worth building, without knowing exactly where that work will take us.

There are still features we want to improve, challenges we need to solve, and people we'd like to reach.

Some days are productive and satisfying. Others involve unexpected problems, difficult decisions, or work that nobody outside our team will ever notice.

And that's simply part of running AICHIKI.

If there's one thing I'd like people to understand, it's that an AI roleplay platform is about much more than artificial intelligence.

It's about the people who develop it, the people who manage it, the people who help maintain its community, and the users who fill it with their own stories and characters.

Every feature, announcement, update, and support response involves work that rarely receives much attention.

And behind all of it are two people trying to balance a business, a family, and a project that has become a significant part of their lives.

We're not a large company, and we don't pretend to have everything figured out.

We're simply building AICHIKI, learning as we go, and trying to make it better with each update.

And I think there's something worth sharing in that.