Question 1

How do we integrate the Evaluation Layer Embeddable UI—iframe vs secure link—and how is access controlled?

Accepted Answer

You can embed the Evaluation Layer via a simple iframe or open it as a secure link, using token-based authentication to gate access. The UI is fully white‑label, so you control the visual layer while CrazyGoldFish handles evaluation, feedback, and storage behind the scenes—no AI team required.

Question 2

What multimodal inputs does the embeddable UI accept for subjective evaluation?

Accepted Answer

The embeddable UI supports text, handwriting scans, diagrams, audio, and video for truly multimodal assessments. Pair submissions with rubrics and/or model answers to drive consistent, explainable scoring and targeted feedback.

Question 3

Does the embeddable UI support standardized IELTS/TOEFL evaluation flows out of the box?

Accepted Answer

Yes. The single‑CTA embeddable UI can run standardized IELTS/TOEFL-style evaluations with rubric‑based scoring, instant feedback, and audit trails—delivering institution‑ready outputs without custom build.

Question 4

What results can we retrieve after an embeddable evaluation session?

Accepted Answer

Use the Final Results API to fetch exam‑level scores, section‑wise summaries, model answers, and step‑wise marking, along with teacher feedback fields (e.g., isApproved). Outputs are available as structured JSON and exportable artifacts (e.g., PDF/CSV) for LMS/ERP dashboards.

Question 5

How are human‑in‑the‑loop review, approvals, and rechecks handled in the embeddable flow?

Accepted Answer

Human‑in‑the‑loop is built in: educators can approve, adjust, or recheck results with full audit logs and version history. This safeguards quality on subjective tasks while preserving traceability for moderation and appeals.

Question 6

Can we brand the embeddable experience to match our product’s look and feel?

Accepted Answer

Yes. The UI is white‑label and supports branding controls like colors and fonts, so the experience feels native while leveraging CrazyGoldFish’s evaluation engine.

Question 7

Is the embeddable evaluation flow aligned with CBSE/ICSE curricula and GDPR requirements?

Accepted Answer

Yes—workflows, rubrics, and reporting are CBSE/ICSE/GDPR aligned, with role‑based access and audit‑ready logs. This lets you deploy standardized, compliant evaluations across regions while maintaining privacy‑first operations.

Question 8

What accuracy should we expect from embeddable evaluations, and how do rubrics/model answers affect it?

Accepted Answer

In practice, teams target up to 95% accuracy by pairing rubric‑aware scoring with model answers for structured tasks. Human‑in‑the‑loop checkpoints handle edge cases, keeping outcomes trustworthy and explainable.

Question 9

Can we start with the embeddable UI and then extend into REST/JSON APIs as we scale?

Accepted Answer

Absolutely. Many teams begin with the white‑label UI for fast deployment, then add REST/JSON APIs to ingest submissions, configure rubrics/model answers, and wire results into personalization and reporting pipelines.

Question 10

What artifacts and analytics are produced from the embeddable evaluation flow?

Accepted Answer

You receive detailed feedback JSON, annotated copies, and section‑wise summaries that highlight strengths and improvements. These standardized outputs feed downstream into action plans, content generation, and exports for SIS/LMS/ERP analytics.

Welcome

Authentication

Evaluation Layer

Personalization Layer

Content Gen Layer- AI Studio

Classroom Observation Layer

Integration Ecosystem

Standardized Language Exams

Embeddable UI for Exam Evaluation

FAQ