Context: January 2026

The end of last year was intense for the whole team. Between our presence at Devconnect with Starknet, starting to search for potential clients to validate the idea, and exploring potential changes to the platform, we've been moving fast.

To be clear about where we are: Kliver understands its role today as a new type of AI roleplay and data annotation platform. Our goal is to build the best infrastructure for creating simulations that reproduce high-ambiguity scenarios — situations designed to train lateral thinking, creativity, empathy, and other social skills.

Product

Our AI roleplay platform — just one of the features we've envisioned for Kliver — is now ready, and we're starting to work with our first users. During January, Kliver will begin being used for consultative sales training, one of the areas where we believe social skills training is essential and where education techniques depend heavily on roleplay.

We're proud of the technical changes that emerged from our evolving understanding of the problem. The platform has mutated significantly, and we've integrated several features we consider differentiators. What remains now is a long road of iteration with clients and UX/UI design improvements.

Clients: Our Focus

As we mentioned in our first post (Genesis Context), our main obsession now is communicating effectively what we consider key and validating our ideas against real clients — both B2B and B2C.

On the B2C side, we've already started the tests mentioned above, which will roll out this month, and we're approaching potential allies we consider strategic.

On the B2B side, we've made progress with contacts at technology companies whose main bottleneck right now is business development. We're building bridges and looking for synergies. We believe it's key to work with clients who truly inspire us — so we can build products they feel confident using internally.

Finally, we're actively seeking connections with AI specialists and professionals — people working at AI labs, AI companies, or experts in the field — to discuss what we believe are critical topics: model collapse and RLHF (Reinforcement Learning from Human Feedback).

We believe now is the time to move forward with experimentation and platform usage.

Next Steps

As part of our progress, we're establishing a clear roadmap for the beginning of the year. We want to define where we're aiming for Q1 and Q2 — which we'll be presenting next week.

We're also organizing internal priorities and the future of the team, searching for advisors in areas we consider key to executing our vision. Right now, we're in discussions to add — either full-time or as advisors — people with experience in Machine Learning (especially NLP, to improve our metrics and discuss specific RL topics) and people with experience in the tech industry.

We hope 2026 will be a year to remember for us.

*Published on Kliver Hub · January 2026*