// a South African firm. Named engagements, running systems, evidence trails. Every promise ends in a command you can run, not an adjective.
/01 practice areas
Each is a named engagement with a delivery shape, a stated dependency on you, and an exit criterion written as a command rather than an adjective.
An agent that does a named job inside your system, on your infrastructure, under your credentials. You buy a running system with an evidence trail, not a capability.
A standing graph of scheduled runs, each with a named output, plus the evidence trail it produces. Not a support retainer and not a seat count.
A commercial process that completes without a human on the success path. That property is machine-asserted, not observed. Most of the design work is deciding what the machine must prove.
Independent attack on agentic systems we did not build. Proposed on internal evidence. It has never been sold, and this site says so rather than implying otherwise.
/02 method
The work is decomposed before anyone writes a line. Each lane carries a named output and an exit criterion. If the criterion cannot be written as a command, the lane is not ready.
Lanes run concurrently against the same contract. Speed comes from isolation, not from hurry. Nothing merges on a promise.
The check is independent of the author, executable, and it refuses rather than warns. A system that grades its own homework is marketing. Ours gets marked by something that wants it to fail.
/03 the gate
One honest footnote. This page is verified by its own maker, which is internal assurance, and we label it as such. Independent assurance is the fourth practice area, and it stays marked proposed until someone has bought it.
/04 boundary
/05 start
One conversation. We map the process, name the output, and write the exit criterion together. If agentic AI is the wrong tool for it, we say so in the same meeting.
labs@potstrategy.com