Category

Safe retries in automation: avoiding duplicate payments

The automation that paid the invoice twice. Why retry is the most dangerous button in automation, and what buyers should demand before they trust a robot with money.

Architecture

Mounika Surla

AI agent guardrails: budgets, grants and prompt-injection defence

Guardrails are not one filter on the model's output. They are a set of limits around the whole run: budgets, grants, schema validation, untrusted-content fencing, a tool-call ledger and evaluations before publish.

VeloPhex Engineering

Bot identity: non-human identity for RPA and AI agents

Robots and AI agents are non-human identities that authenticate thousands of times a day. Give them key pairs, short-lived tokens, narrow grants and outbound-only connections, not a password that lives forever.

VeloPhex Security

Scaling RPA bots: what actually breaks as the fleet grows

Going from fifty robots to a thousand rarely fails where people expect. It fails in job claiming, presence, retries and reconnect storms, and AI agents add new limits of their own. Here is what to design for.

Engineering Enterprise Automation Python Architecture AI & Agents Product