24 December, 2025 - Last week in TI
- Rahul Subramaniam
- Releases
- December 24, 2025
EduPaid v2.12.0: Comprehensive Webhook Integration System
Real-time webhook notifications for subscription, payment, and refund events enable seamless integration with external systems and automated workflows for providers.
- Real-time event notifications with flexible configuration: Subscribe to subscription status changes, payment completions/failures, and refund completions through comprehensive webhook system with rich payloads including event details, provider and student information, subscription status, payment amounts, and refund details.
- Secure endpoint management with robust delivery: Create, update, activate, and deactivate webhook endpoints through secure API with per-endpoint event type filtering, ensuring webhooks are never lost or duplicated and enabling confidence in critical business integrations without requiring database access.
LearnLens Studio: Adaptive Pre-Test & Streamlined Learning Flow
Your learning journey now begins with intelligent diagnostics that ensure courses are perfectly calibrated to your knowledge level, plus streamlined controls that keep you in the flow state.
- Adaptive Pre-Test with ZPD Analysis – Real-time diagnostic engine analyzes your understanding and visualizes your knowledge landscape: Known Concepts, Concepts to Learn, and your personalized Recommended Learning Path. Select or deselect topics to learn exactly what you need—no wasted time on mastered material.
- Auto-Play Next & Skip to Quiz – Stay in the learning zone without interruption. Videos automatically advance after quizzes to maintain momentum, or skip directly to assessments if you’re confident. Competency-based progression that respects your time and prior knowledge.
- Quiz Accuracy Guarantee – New backend validation ensures every multiple-choice question matches its explanation perfectly, eliminating answer mismatches. Plus, smart randomization prevents pattern guessing across all assessments.
- Key Improvements – Fixed stale questions in practice quizzes, smart completion flow to Mastery Challenge, precise note timestamps, and reliable segment-to-quiz transitions.
Athena Applets: Strong Quality Metrics & Multi-Grade Expansion
8 new lessons generated, 9 lessons approved this week, achieving a 55.6% first-time pass rate. The review system demonstrates strong efficiency with detailed feedback analytics across multiple grade levels, while the review queue continues to grow with 90 lessons across Grades 4-6.
- Production output – 8 new lessons generated and 9 lessons approved this week.
- Quality improvement – 55.6% first-time pass rate.
- Efficient iteration cycle – 23 total iterations completed in the past 7 days (14 feedback rounds, 9 approvals) with an average of 1.4 comments per lesson.
- Multi-grade coverage – Lessons approved across multiple grades: Grade 4 (1 lesson), Grade 5 (3 lessons), Grade 6 (5 lessons), with a robust review queue of 90 lessons ready for validation (Grade 4: 23, Grade 5: 5, Grade 6: 62).
- Complete lesson catalog – See the full list of all uploaded lessons across grades in this lesson catalog with direct links to each lesson (95 Grade 3, 3 Grade 4, 4 Grade 5, 8 Grade 6).
BrainTrust: Exclusive Agent Access & Simplified Configuration
Take control of who can hire your AI agents and reduce setup friction with two powerful new features that make agent deployment faster and more secure.
- Agent Whitelisting – Control access to your agents by whitelisting specific email addresses or entire domains. Make your agents exclusive to your organization, team, or select partners—perfect for internal tools, beta programs, or premium offerings that require controlled access.
- Agent-Level Variables – Pre-configure agent settings to reduce the complexity for users during hiring. By setting default values at the agent level, you minimize configuration steps and eliminate friction in the hiring process, making it easier for non-technical users to get started with your agents instantly.
Athena Accelerator: Repair & Redirect Interventions (Beta) — Completing the Adaptive Recovery Stack
This release completes Athena’s three-level intervention ladder with the Beta launch of Repair Mode and Redirect recommendations. When scaffolding alone isn’t enough, the system now intelligently escalates to focused corrective drills that target persistent misconceptions at the skill level—not just the current question. If deeper gaps are detected, students are gently guided toward supplementary resources, ensuring no learner gets stuck in an unproductive grind.
- Repair Mode Interventions – When a student shows repeated struggle with the same misconception across multiple questions, the system now pauses normal practice and deploys a short, targeted repair drill. These rapid-fire sequences (3–5 questions) are designed to fix the underlying rule or representation—like gym equipment for the mind—so students stop making the same class of error going forward.
- Redirect Recommendations – If repair drill responses indicate continued struggle or a potential prerequisite gap, students receive a non-enforced recommendation to visit passive learning resources. Currently, this links to Athena articles and videos; future releases will enable in-platform redirection to prerequisite skills when gaps are detected.
- Unit Coverage – Repair and Redirect interventions are now available across the first four units of Grade 3, matching our scaffolding coverage and enabling a complete adaptive experience for early adopters.
- Intelligent Escalation Logic – The system uses behavioral signals—repeated same-misconception errors, low learning yield during scaffolds, and answer reveal patterns—to determine when escalation from scaffold to repair is warranted, ensuring interventions ‘pay rent’ by reducing total time-to-mastery.
- Enhanced Journey Visualization & Dev Tools – The student journey graph now captures the full intervention ladder—scaffolding, repair, and redirect events—providing complete visibility into each learner’s adaptive path. Stakeholders can enable Dev Mode on the site to rapidly preview and test different repair sets for any skill, accelerating content QA and iteration.
- Known Beta Limitations – As with scaffolding, you may encounter occasional cases where correct student input doesn’t match expected answer formats, or where repair questions feel redundant. Similar edge cases from the scaffolding release may also appear in repair mode. Your feedback continues to be invaluable for system refinement.