SVEED turns review of already-taught content into a short, guided, offline-first routine — designed to keep working when connectivity is limited or academic support at home is unavailable.
View White PaperSchool teaches. SVEED supports the review.
Learning has to continue after class. Homework and other forms of at-home review have long helped students practice and consolidate what was taught in school. The challenge is that the conditions needed to sustain that continuity are not equally available across households.
International organizations recognize multiple dimensions of education's last mile. The OECD notes that it can extend into students' homes and personal devices Source: OECD 2023 ; UNICEF uses the term last-mile learners for children who are being left behind and are among the hardest to reach Source: UNICEF 2022 .
A teacher leads an entire class. After the lesson, each child may need to revisit different points and consolidate a different part of what was taught.
Homework is a long-standing instructional practice. OECD evidence indicates that short, regular assignments can support learning and self-regulation Source: PISA 2022 .
Time, quiet space, connectivity, devices, and academic support are not equally available. OECD data links these differences to socioeconomic gaps in homework participation Source: PISA 2018 . SVEED was designed to reduce some of these dependencies.
SVEED is a lightweight Android-based system for continuity of learning. It targets a specific expression of education's last mile: the gap between content taught collectively in the classroom and the individual review and consolidation that happens afterward. Guided curriculum review is the mechanism SVEED uses to reduce that gap.
The starting point is a curriculum skill already covered at school, identified by the child or selected through teacher mode.
After an incorrect answer, the child can receive a hint and try again. Support guides the reasoning process without turning the session into a performance scoreboard.
When difficulty persists, predefined rules can step back to an associated earlier skill and then return the child to the original learning path.
School teaches. SVEED supports the review. The product is not intended to replace teachers, classroom instruction, or the curriculum, and it should not autonomously introduce curriculum content that has not yet been taught.
A short, predictable, individual routine designed to help the child keep going when difficulty appears.
The core of SVEED is a daily review session of approximately 15 minutes. The interface stays simple; the complexity remains in the logic that organizes the sequence, offers support, and adjusts difficulty.
In the flow without teacher input, Nexo Visual (visual topic confirmation) helps the child identify the topic covered at school before starting the review.
The system presents locally generated exercises within a closed curriculum universe, keeping the session close to the selected skill.
When an answer is incorrect and a hint is available, the child can ask for help and review a prompt before trying again.
The second chance lets the child resume the reasoning process. The Concept Library can explain a related concept separately from the hint for that specific exercise.
Orbit rules can temporarily reduce difficulty or recover a prerequisite, always aiming to guide the child back to the target skill.
Principle: an error is a signal to provide support and reorganize the experience, not a mechanism for punishment or competition.
Curriculum content, graphs, templates, and rules remain on the device, reducing network dependence during review.
The operating system's text-to-speech can read prompts and instructions, reducing dependence on an adult reader.
The orbit uses TARGET, TARGET-1, TARGET-2, and TARGET+1. The curriculum graph is used to recover prerequisites, not to advance into untaught content.
SVEED is a functional product at a pre-external-validation stage. The product and its underlying mechanisms can already be examined; the next maturity step is demonstrating behavior outside the environment close to product development.
This section presents the project's current state. It does not use simulated adoption, user, or impact figures.
Automated tests
145
0 failures · 0 errors · 0 skips after the Evaluation Gate
Curriculum structure
6 × 9
6 subjects · Grades 1–9 · curation at different stages
Formal external validation
Pending
Next step: evaluation and a documented pilot
Release builds compile successfully. Final institutional signing and packaging are still being completed.
The project still needs to observe autonomy, routine continuity, hint use, and outcomes compatible with the review scope in an external context.
Current position: a functional educational product with architecture sufficiently developed to justify evaluation and a pilot, but without a formal external outcomes study yet.
SVEED is designed to complement the school's work and increase the child's autonomy at home, without requiring an adult to guide every exercise.
In teacher mode, the school can indicate the curriculum skill that should guide that day's review at home.
The interface combines optional audio, hints, a second chance, and a short routine to support progressively independent use.
The parent panel provides a simple session-level summary, including time, subject, exercises, errors, and hint use.
The offline-first architecture and low technical execution cost support deployment scenarios on Android devices and in limited-connectivity contexts.
The goal is to add a layer of individual review after class, not to create a parallel school inside the device.
SVEED describes security according to what is implemented today, without turning technical intent into a promise.
The review does not require the child's name, national ID number, email, or phone number in order to work.
Persistent technical identifiers and session records exist. Therefore, the accurate description is not absolute anonymity.
The educational session runs locally and does not depend on continuous connectivity to cloud services.
Questions shown to the child are produced from local rules and templates, not from calls to generative models.
The current implementation records session-level information such as time, subject, skill, correct answers, errors, and hint use.
The product does not currently claim to be tamper-proof, provide absolute anonymity, or guarantee cryptographic integrity against intentional tampering.
SVEED is designed for guided individual curriculum review. In its current state, it should not be presented as a system that:
Technical-institutional document covering the educational problem, pedagogical flow, architecture, current security limits, maturity stage, and next evidence steps.
PT
EN
ES
