FinArb

Their pitch is 'AI finds market inefficiencies for you' — which is adorable, since the inefficiencies that survive long enough to be packaged into a SaaS product are the ones that weren't worth exploiting in the first place. Meanwhile, Claude and GPT-4 are teaching 19-year-olds to code quantitative strategies that would have required a Goldman internship five years ago.

Quant researchers, strategy developers, and risk analysts — the entire value chain that justifies charging retail traders for 'institutional-grade algorithms' — are becoming weekend projects for anyone with Cursor and a decent understanding of Python. The humans they employ are now competing with an AI that doesn't sleep, doesn't ask for equity, and doesn't need Series A funding to improve.

Their moat is 'proprietary algorithms' in a market where the barriers to algorithmic development just collapsed to the cost of an OpenAI subscription. The real kicker: the algorithms that actually work don't get democratized — they get deployed at scale by the people who found them.

Website unavailable for analysis, but the domain suggests they're positioning as an AI company rather than adapting to AI — which in fintech usually means they're selling the sizzle while the steak gets commoditized by foundation models that don't need a business plan.

Making money from market inefficiencies survives; paying a middleman to find them for you doesn't. AI democratized the arbitrage discovery, which paradoxically killed the arbitrage business — everyone having the same tools means no one has an edge worth paying for.