AI Generation

Swiftscore uses AI to do the heavy lifting across the evaluation lifecycle — scoring observations, summarizing conversations, synthesizing trends across your school, and answering questions through your built-in assistant. This article gives you an overview of where AI generation shows up across the product and points you to the deeper guide for each one.

Evaluation scoring

When you generate an evaluation, Swiftscore scores every component of your framework, writes a narrative for each one, and cites the evidence behind the rating. This is the core of what makes an AI-generated evaluation different from a blank form — every score is meant to be traceable back to something that actually happened during the observation.

While an evaluation is generating, you'll see live progress as Swiftscore works through your framework — a sign that scoring is actively underway rather than a static loading spinner.

For the full mechanics of how scoring works — including how scores roll up into domain and overall ratings, how calibration shapes the language, and what happens if something goes wrong mid-generation — see Evaluation Scoring.

Conversation summaries

After a debrief conversation with a teacher, Swiftscore can generate a structured summary — what was discussed, what was agreed, and what the next steps are — directly from the conversation. This saves you from writing up notes from scratch after every conversation, and gives both you and the teacher a shared, accurate record to refer back to.

You can edit the generated summary before finalizing it, and regenerate it if you want a different take on the same conversation.

Professional development reports

Beyond individual evaluations, Swiftscore can synthesize patterns across a school's recent evaluation history into a single Professional Development (PD) report — recurring themes, trends, and recommended next steps for your staff, grounded in your teachers' actual evaluation data rather than generic suggestions.

This is useful when you want a school-level view of how your staff is trending, rather than reviewing evaluations one at a time.

See Generating a Professional Development Report for Your School for a full walkthrough.

Edith, your AI assistant

Edith is Swiftscore's built-in AI assistant, available from a chat panel anywhere in the app. She can answer questions, help you think through feedback, and pull in relevant context from your account as you work — with her responses streaming in as she generates them, rather than appearing all at once.

For everything on opening Edith, chat history, and how she responds to context, see Settings, Edith, and Notifications.

How AI fits into your workflow

AI generation in Swiftscore is meant to accelerate your work, not replace your judgment. Generated content — scores, summaries, and reports — is always editable. You can review, adjust, or regenerate any AI output before it's finalized, and nothing generated by AI is treated as final until you say so.