Format
7 articles
CTO's PerspectiveAI & Automation
Companies that adopt AI coding agents open 35.7% more pull requests a year later, and each pull request takes 56.1% longer to review. Finished work moved much less.
Oleksandr Veklych8 min read

CTO's PerspectiveProduct & Delivery
Use a two-bottle order to test duplicate payment notifications, lost warehouse replies and recovery without issuing a second shipment for the same purchase.
Oleksandr Veklych8 min read

CTO's PerspectiveSecurity & Reliability
Follow one export through an API, a worker and a download link to check whether a customer can retrieve another tenant’s records.
Oleksandr Veklych8 min read

CTO's PerspectiveProduct & Delivery
A subscription cancellation example shows how teams can assign review ownership and test a coding agent’s pull request against an independent product rule.
Oleksandr Veklych8 min read

CTO's PerspectiveAI & Automation
A field service search pilot should test asset identity, manual applicability and source verification before asking technicians to rely on a generated answer.
Oleksandr Veklych8 min read

CTO's PerspectiveAI & Automation
A retail support pilot should measure verified resolution, repeat contacts and how much work an adviser must do after a customer leaves the chatbot.
Oleksandr Veklych8 min read

CTO's PerspectiveAI & Automation
A practical release review for a finance agent, from the cases used in evaluation to the operator who can stop an incorrect action.
Oleksandr Veklych8 min read