The industry default is the platform: move your assessments into it, retrain your staff on it, and rent access to your own work forever after. We built the opposite, on purpose.
Software that disappears into a workflow people already have tends to get used. Software that asks the workflow to reorganise around it tends to get resented, then worked around. That shapes every integration decision in Apollo.
Apollo reads the files your students already submit, in the formats your templates already use, from the folders your trainers already keep. Nothing is rebuilt as quiz questions, nothing is migrated, nobody spends a term learning a new system. If your assessment workflow works, it should survive the arrival of AI untouched.
Marking runs on your organisation’s own Anthropic account, billed at cost by them, with the spend visible in your own console under a cap you set. We add no margin to tokens and sit in none of the traffic. The AI relationship belongs to your RTO, which means you can read it, cap it, or end it without asking us.
Software should fit the organisation that already exists.
Integration has an exit condition. Apollo works beside your SMS and LMS rather than inside them, holds none of your records, and stopping is an uninstall rather than a data-extraction project. A tool you can leave freely is a tool you can adopt calmly, which is the posture an RTO should want with any AI it lets near assessment.