sjt-at-rev

3 ms·

1.
Agent to stress-test Wall Street analyst reports: Chipotle (CMG) Example(static-www.revelata.com)
3 points by sjt-at-rev 6mo ago|1 comments
2.
by sjt-at-rev 6mo ago
Used Claude + deepKPI to break down the arguments in a sell-side report, pull more detailed data to support them, and surface data that contradicts them. Demo taking a look at RBC's Chipotle note. Every analyst emphasizes the numbers t
3.
by sjt-at-rev 6mo ago
Yeah, I've had good experience with Kimi too. Good for the price point, for sure. Anthropic models are still the best for me -- as long as you don't ask them to do something they don't want to -- but also, way too expensive f
4.
by sjt-at-rev 6mo ago
I think the other side of that coin is how much effort it takes to get it to do what you need. Our pipeline is a sequence of very precise tasks where subtle contextual cues matter a lot, and there are a large classes of related error modes.
5.
by sjt-at-rev 6mo ago
We've been testing it extensively and its performance is no better than prior versions, and in many cases worse than open weight models like GLM. Gemini3.1 Pro is so significantly better. To me, the play is: open weight on a provider l
6.
GPT 5.4 in practice – Stinks?
9 points by sjt-at-rev 6mo ago|14 comments
7.
API and MCP that pumps data from SEC filings into your agent context(medium.com)
2 points by sjt-at-rev 6mo ago|1 comments
8.
by sjt-at-rev 6mo ago
Pulling KPIs from SEC filings is trickier and more error prone than it looks at first blush, and the risk of plausible land-mine hallucinations is still a deal breaker for a lot of serious finance folks. Our solution is a carefully engineer