Taimi - The AI Market Intelligence
Look! A seedling - or should I say “taimi” in Finnish - is growing from a tree stump! But what is it actually? Taimi is a simple tool that tries to give up-to-date information about the license prices of most used Agentic AI development tools that work with existing codebases. At the moment we provide a html page and an API for anyone for free to use.
See for yourself at taimi.market.
Proactive FinOps
At the moment, the AI development tool landscape is moving fast. While it is important to evaluate the general suitability of the tool to a current task or for an organisation, we have noticed that increasing the visibility of cost of these tools is becoming ever more important. There’s real danger that costs could spiral out of hand. While FinOps FOCUS specification tries to answer “What did this cost us?”, Taimi tries to answer “What will this cost us?”.
Shared GreenOps
At this point, you might be asking what is powering Taimi? Well you guessed right, it’s AI. Then you might ask, isn’t using AI bad from GreenOps perspective? Short answer: yes and no. While the mantra of cost being a good proxy for sustainability might be a bit distant in this specific case, the initial vision of Taimi was this: what if we could generate this information once and provide it for free for anyone to reuse so that no one else needs to generate it?
AI philosophy
First and foremost: an experiment. We at Kanto realised that there’s a use case for AI: the price information location and content is volatile due to the fast speed of the market. No amount of manual work could keep up the pace of updating the data endpoints. We first started lazily trying to achieve as much as possible with AI alone. Then we realized that adding more determinism with scripts, templates, and other guardrails could be beneficial. But this is just a basic iterative thing to do in software development.
Governing the non-deterministic
Whether or not you should trust the non-deterministic AI generated data is up to you. We at Kanto are focused on these questions: What could be removed and what should not be built in the first place? Then there’s this: How do we design and build systems that are safe, cost effective, and sustainable? Determinism is still needed, that’s the actual engineering part. We’re just starting to realize the actual cost of AI for all of us. But you can be damn sure we’re watching.
