Tag
11 posts tagged.
A launch handoff for a client's AI app is two layers, not one: the classic software layer (code, environments, credentials, tests) plus the AI layer (provider account and keys, prompt library, evaluation set and baseline, cost monitoring, model version and deprecation plan) and a one-page runbook.
How agencies turn a client's vague 'make it accurate' into a measurable, defensible AI accuracy target in 2026: a target, a floor, the right metric per task, and a demo that cannot blow up.
The requirements-gathering questions software agencies must ask before quoting an AI app build in 2026, grouped into eight sections, plus the six answers that should move your price.
Split every AI app build into deterministic scaffolding, eval-gated AI R&D, and a non-determinism contingency, then estimate and bill each layer differently.
The four models agencies use to bill clients for ongoing AI/LLM API costs, with a worked monthly-bill example and a decision matrix for choosing one.
Generic white label templates go silent on the clauses that carry the real risk in AI app builds. Here are the eight your reseller contract needs, plus a copy-paste skeleton.
The generic retainer templates on Google were built for lawyers, not software agencies. Here is a copy-paste retainer for AI app work, with the six clauses to add and a utilization-honest way to price the monthly fee.
A copy-paste change request template for fixed-price AI app builds, plus five AI-native clauses and a simple way to price every scope change so it stays profitable.
A complete, copy-paste statement of work template, plus the five AI-native clauses generic SoW templates leave out: platform, IP and source ownership, migration, capped revision rounds, and hosting plus data residency.
