Sabitlenmiş Tweet

We're working on something a bit unusual. This is not a typical hackathon project and definitely not another “AI trading bot.” The idea is to build an autonomous financial agent that can distinguish signal from noise, understand market regimes, make decisions under constraints, manage risk, execute trades, and explain its own actions. For me, this is not about winning anything; it is a capacity test to see whether we can already build something that behaves like a real system rather than just a predictive model.
We are not starting from zero. There is already a planning framework (PCHS) that governs how decisions are made, a signal–noise–regime perspective for interpreting markets, a structured understanding of quant strategies, an experimental hybrid GRU/MACD model, and ongoing research work. On the infrastructure side, there is access to a DGX / Blackwell-class local compute environment, and there is also prior proof of delivery through PulsarWave, which reached the Top 6 at the Google Vertex AI Hackathon.
The way we work is also intentional. We use Kuwaki as our core knowledge system for structured thinking, research, and documenting decisions, and we use Discord for coordination and fast iteration. We are starting this solo by default and will only add someone if that person can contribute immediately, fully own a part of the system, and operate without onboarding. If this resonates, send what you have built, what you can own, and how you think about signal versus noise.
Apply now!!!: info@tensibility.ai
Project: lablab.ai/ai-hackathons/…
Our Team Brief: drive.google.com/file/d/1-8llqP…

English







