A change is picked up from the tracker, classified, and routed to the right lane. Impact is computed deterministically from a knowledge graph, a risk-ranked regression suite is selected with a reason for every include and exclude, the tests run against the banking APIs, defects are filed with evidence, and every step is traceable end to end. Pick a surface to begin.
Watch the agents work in real time — intake, impact, test design, execution, the human approval gates, defects and the evidence report.
The native graph of requirements, test cases, APIs, code modules and defects. Select a change and the impacted subgraph lights up with the ranked regression suite.
The AlAhli Online and Cards Management APIs the tests run against — try the endpoints and see the daily-limit and credit-limit controls the agents verify.
CR Impact lane. Graph traversal finds exactly which requirements and test cases the change touches, a risk-ranked regression suite is selected, the 11-agent SIT runs the impacted tests, and the run is written back to the graph as an audit node.
New Request lane. A net-new feature with no existing coverage — the agent generates user stories and test scenarios and gives an indicative, parametric effort estimate with its basis. Ready for SIT once built.
Standard SIT. An ordinary requirement runs the full 11-agent pipeline unchanged — design, review gate, execution, defect gate, evidence — exactly as delivered.