One-to-One Learning Platform · Student / Teacher / Operations
From manual matching to a scalable learning service
Nine connected service stages take a learner from the first request to ongoing lessons. AI supports teacher matching, lesson planning, and feedback, turning work once held in individual staff members' heads into a repeatable operating model.
The ChallengeOne-to-one learning takes more than
One-to-one learning takes more than
finding a teacher and starting a lesson
Behind every lesson is a chain of conversations, teacher selection, trial bookings, purchases, planning, delivery, and feedback. Students, teachers, and operations teams all need to stay in sync.
Student schedule.xlsxLocal · manually maintained
| Student | Teacher | Time | Status |
|---|---|---|---|
| Student Wang | Teacher Li | Sat 18:00 | Pending |
| Student Zhang | ？ | ？ | No match yet |
| Student Liu | Teacher Chen | Sun 10:00 | Trial complete |
| Student Zhou | Teacher Wang | ？ | Not asked |
| Student Sun | ？ | Wed 19:00 | Awaiting reply |
ZZhang's parent12
When can the teacher confirm the trial?Is the teacher we discussed still available?My child is free this weekend.
LTeacher Li5
Can you send me the student's background?Which chat has their requirements?
Student Wang · trial Sat 18:00
Remind Teacher Li
Remind Teacher Li
Teacher Chen is unavailable Wednesday
Reschedule
Reschedule
Student Zhao's budget
Up to 200
Up to 200
Teacher LiChildren's English · weekday evenings
Requirements, matching, trials, and progress lived across chats, spreadsheets, and memory.
As student and teacher numbers grew, that manual approach stopped scaling. Taoli Vision needed a better way to run the service: How could a people-intensive service become a connected system that moves each learner forward and scales with the business?
Where the Process Broke DownWhen context gets lost,
When context gets lost,
the whole service feels the impact
The friction surfaced across intake, matching, trials, and follow-up in four recurring ways.
One clear service state for every learner
GeekOnUp turned fragmented conversations into a visible workflow that teams can advance, hand over, and improve together.
01 / 09 · Scroll to progressStudentAITeacher
01Share learning goalsStudentNot submitted → Ready to match
A guided intake captures the learner's background, goals, budget, and schedule, turning an open-ended request into information the team can use.
Challenge One
Students may not know how to describe what they need
Students' first requests rarely contain enough detail for a confident match.
A useful recommendation depends on current ability, learning goals, budget, availability, and teacher preferences.
We designed a guided intake to capture that context at the start.
GGuided intakeIn progress
Learning briefReady for matching
- Learning goalChildren's spoken English
- Current abilityGood listening · speaking needs practice
- Lesson frequencyTwice a week
- AvailabilityWeekdays after 20:00
- BudgetCNY 200–300 per lesson
- Teacher preferencesExperience teaching children
Once intake is complete
Automatically generated
Automatically generated
More teachers made manual matching harder
Teachers in the same subject may specialize in children's English, adult learning, conversation, or exam preparation.
Their teaching styles, ideal students, and schedules differ too.
We applied AI to the high-volume first pass of teacher selection.
AITeacher matchingReading learning brief…
Matching criteria · from the learning briefChildren's spoken EnglishWeekdays after 20:00CNY 200–300 per lessonExperience teaching children
24teachers reviewed
GGuo
ExamsIELTS
WeekendsSSun
ChildrenPicture books
WeekendsZZhang
TeensPrimary school
WeekendsLLi
ChildrenPhonics
Weekday eveningsHHu
SpeakingEveryday topics
Weekday eveningsXXie
AcademicWriting
Weekday eveningsLLiu
TeensMiddle school
Weekday eveningsWWang
ChildrenEarly speaking
Weekday eveningsLLin
SpeakingTravel
WeekendsLLuo
ExamsTOEFL
WeekendsCChen
ChildrenSituational speaking
Weekday eveningsTTang
AcademicResearch
WeekendsZZhao
ChildrenIntroductory grammar
Weekday daytimeZZhou
TeensPhonics
WeekendsHHe
SpeakingFilm dialogue
Weekday eveningsZZheng
ChildrenGraded reading
Weekday eveningsLLiang
ExamsCollege English tests
Weekday eveningsMMa
TeensGraded reading
Weekday eveningsXXu
AcademicWriting support
Weekday daytimeWWu
ChildrenPlayful conversation
Weekday eveningsGGao
SpeakingPronunciation
WeekendsXXu
ChildrenEnglish expression
WeekendsSSong
ExamsGraduate exams
WeekendsZZhu
TeensInterest-led learning
WeekendsThree AI-assisted recommendations · with clear reasoning
- 1Teacher LiEight years teaching children; strong at interactive, confidence-building lessons
- 2Teacher ChenHelped 40+ young learners build speaking confidence from the start
- 3Teacher WuConsistent weekday availability after 20:00
Connected Delivery
The journey continues after the match
A recommendation alone does not make a complete service. The platform carries context through the trial, purchase, course plan, lessons, and feedback after a teacher is selected.
Student
Book a trial and purchase lessons,
then follow the plan and learning progress.
Teacher
Confirm schedules, review learner needs, plan lessons, and submit feedback.
Operations
See where each learner stands and step in when personal support is needed.
A shared status now answers questions that once required searching through chat history. The operations workspace becomes a way to keep the service moving, not just record what happened.
Practical AI
Use AI where repetitive work slows the team down
Beyond matching, AI helps draft course plans and post-lesson feedback, two tasks teachers handle repeatedly.
Course plan · Student Wang · children's EnglishReview requirements
1Read the learning briefTurn the learner's background, goals, budget, and availability into usable inputs.
2Draft the course planOutline learning stages, priorities, and lesson structure in seconds.
3Teacher reviews and editsThe teacher adjusts pace and priorities, keeping professional judgment in the process.
4Share with the learnerApproved plans and feedback appear in the learner's course view.
Less time drafting from scratch, with teacher expertise where it matters. Feedback and assignments follow the same pattern: AI drafts, and the teacher reviews before sharing.
Three workspaces, one learner journey
The student app, teacher workspace, and operations console work from the same learner record and the same service journey.
Teacher workspace: start with the full learner context
- 1Review learner details and the structured brief without asking again
- 2Confirm trials and manage upcoming tasks
- 3Refine an AI-assisted course plan
- 4Teach, then share feedback and assignments
Learners avoid repeating themselves, teachers get the context they need, and operations teams see progress in one place.
Built to Learn
Teacher profiles
Improve with every real lesson
Initial matches use teacher experience, specialties, teaching style, and availability. As lessons happen, trial outcomes, purchases, completion, and long-term feedback add evidence to improve future decisions.
Product Screens
The experience in practice32
Actionable Data
See exactly where the learner journey slows down
With every key action in the platform, the team can see the full conversion path and improve the service before the final outcome.
Illustrative journey; figures do not represent actual business results.
Once the bottlenecks became visible, the operating metrics moved.
Business Impact
A service built to scale beyond manual coordination
Connecting service stages and applying AI to matching, planning, and feedback delivered measurable improvements after launch.
Immediate improvement0%↓ Roughly half the wait
Average teacher matching time reduced by
AI handles the broad first pass, leaving exceptions and final decisions to the team.
- 0%Trial conversion↑ Increase
- 0%Learner requests managed
per operations team member↑ Increase - 0%Timely teacher confirmations
and post-lesson feedback↑ Increase
Growing together
A service once run by hand
became a repeatable
operating model
Intake, matching, trials, purchases, lessons, and feedback now form one connected journey. GeekOnUp helped Taoli Vision turn individual know-how into a product capability the business can build on over time.
- Staff track teachers through chats and spreadsheets→One connected journey from intake to feedback
- A process dependent on individual experience→An operating capability the platform can repeat
- Growth requires more manual coordination→The service can scale without adding staff at the same rate