Where it fits
Built for decisions like these.
The work is the same across domains: a high-stakes question, real data, and a conclusion someone has to defend. The domain changes the evidence, not the standard.
Investment research
Macro, rates, and equity questions that need a defensible answer, not a faster draft.
Is the common-factor story real? What does the positioning data actually imply? Which of these three theses survives the evidence?
Trading infrastructure
Latency, throughput, and market-data architecture, measured and recorded rather than asserted.
Where is the real bottleneck? Did the optimization help? What does the measurement actually prove?
Commodity trading
Physical flows, basis, and policy transmission tied to a testable price view.
Does the flow link survive point-in-time data? What would kill the thesis? What is priced in?
Corporate strategy
Market entry, competitive position, and capital allocation under uncertainty.
Is this market real? Who is the buyer? What would have to be true for this to work?
Risk and compliance
Evidence trails, approval records, and data boundaries your risk function can inspect.
Can we reconstruct the decision? Which sources were used? Who signed it?
Data and engineering
Agent workflows, evaluation harnesses, and the plumbing that makes them repeatable.
Does the agent actually work outside the demo? How do we evaluate it? What is the failure mode?
The work we run
Three shapes, one standard.
-
01
A deep researched memo
A single name or question, taken far enough that an analyst can review and a portfolio manager can use.
-
02
Frequent company coverage
Coverage that has to happen often, on a book that is larger than the team can staff by hand.
-
03
A custom programme
Bespoke data loads, analysis, and research-idea finding, inside the same controls.
Later jobs can use what earlier jobs learned, inside the same controls. Scope is explicit.
Who it is not for
A tool-only buyer will be disappointed.
- Teams that want a better chatbot, and no one accountable for the answer.
- Teams that cannot give access to a real decision or real data.
- Teams that want a fixed answer without participating in the verification.
Start here
Bring one decision.
A 2 to 4 week sprint on a live decision. You get the decision record, the verification trail, and the method your team can run next.