Reporting and AI Support for an AI Reading Tutor
Reporting for teachers, parents and districts, and Amirabot, an AI support assistant, built for Amira, the AI reading tutor used by over a million students.
- Client
- Amira Learning
- Industry
- Education Technology
- Timeline
- 2022 to 2026
- Platforms
- Web
The brief
What Amira Learning needed
Amira Learning is an AI-powered tutoring platform used by over a million K-5 students. Students are assessed and tutored through Amira, and teachers, parents and district administrators use the results to decide who needs help and with what.
Amira first brought VantaSoft in because it wanted far more teacher reporting than its internal team could deliver on schedule. We worked inside that team to ship it.
The needs grew from there: reports for parents and district administrators, report calculations moved off each teacher's browser and onto the server, and one AI assistant to answer educators' support questions in chat and by email, instead of a separate bot inside each support tool.
What we built
What we built for Amira Learning
-
Reading Rope report
Each student's current mastery across vocabulary, background knowledge, phonological awareness, decoding and high-frequency words, with what to work on next.
-
Class Progress report
Reading mastery for a whole class across the fall, winter and spring benchmark windows, measured against national norms. A second view, Thread Overview, shows each student's percentile on seven reading measures in one color-coded grid.
-
Tracking report
Tutoring time and stories read per student each week, with assessment status and controls to assign benchmark and progress-monitoring assessments.
-
Badges and certificates
A badging system for reading achievements, in English and Spanish. Teachers see each student's usage, award the badges they have earned and print certificates.
-
Parent letters and reports
Dyslexia-risk letters to parents fill in each student's scores automatically, and parent reports show every benchmark window of the year. Both print in English or Spanish, regardless of the assessment's language.
-
Family portal in Hindi
Families in India sign up with a phone number instead of an email address, and sign-up, sign-in and the main report are in Hindi.
-
Server-side reporting
Report calculations moved from the teacher's browser to server-side services, which now feed the benchmark, Class Progress, Reading Rope and progress reports.
-
District and school reports
Tracking, growth, screening and Science of Reading reports for administrators, embedded in the educator app and scoped to each district and school.
-
Amirabot in chat
An AI support assistant inside the educator app. It answers common questions straight away, sorts each conversation by topic and hands off to a live agent during support hours.
-
Amirabot by email
Emails to the support queue get an AI-written reply on the case, or a draft in the support team's inbox. When a person should answer, Amirabot routes the email to the support team instead.
-
Support insights dashboard
Support conversations rolled up by category every night, with a product-issues view and exports, so Amira's product team can see what is driving support volume.
-
Customer success sync
A daily job that matches each account in Amira's customer success platform to its district or school, and keeps roster counts and screening dates up to date.
Results
What it changed
-
Reporting at the depth teachers need
Teachers see each student's mastery skill by skill, the whole class across benchmark windows against national norms, and weekly usage, and they assign assessments from the same place.
-
Four years inside Amira's team
We worked as part of Amira's own development team from 2022 to 2026, on its codebase and to its milestones.
-
A sturdier platform
Beyond new features, we modernized parts of Amira's infrastructure for better performance and stability.
-
An award-winning platform
During our work together, Amira was named a Forbes AI 50 company (2022) and won SIIA CODiE Best Science of Reading Solution and EdTech Awards Best AI Solution (2023).
In their words
“In our experience, development partners often have strengths, but those positives are obfuscated by a fatal flaw. We've worked with teams that create quality code but are too expensive or too slow. We've found folks who are professional but struggle to grasp what we need. Finding a partner that delivers across the board has been a big advantage for us, and I'm very pleased to recommend VantaSoft to you.”
Stack
- React
- Python
- Node
- GraphQL
- AWS
- DynamoDB
- Domo
- Salesforce
- OpenAI
What we did
- Custom Software
- Web Applications
- Integrations
- LLM Integration
Tell us whatyou’re building.
Senior engineers and fractional CTO leadership, onshore in Irvine. It starts with a conversation.
Updated