EdTech platform with live learning, mentoring, and booking workflows
Learning programs, live-session scheduling, mentor bookings, payments, and progress tracking were brought together into a single platform. During the first 8–10 weeks after launch, the product achieved a 2.8–4.6% registration conversion rate. Between 35–48% of new users reached their first value milestone (starting a track or booking a live session), while 6–10% completed their first payment.
About the Project
A web platform for learning that centers on live sessions and mentoring, where the core value is generated through participation and personal feedback rather than passive content consumption. A user picks a track, gets a clear progress path, books live sessions, and — when needed — books additional mentor consultations, accesses materials, and tracks progress. Payments and access rights are implemented directly into the in-product flows, so monetization is anchored to real user actions inside the platform rather than to landing-page promises.
About Customer
Two co-founders who started as instructors and built the platform on their own savings — which made fast hypothesis validation and tight cost control non-negotiable. They came in with a clear idea: launch an online school where users consistently reach the first session, see real learning value, and pay for the next step on their own initiative — without manual sales pressure from the founders.
Greenfield product
No existing software, no off-the-shelf reference — every piece of platform logic, role design, and operational rule had to be defined from zero.
Bootstrapped budget
The architecture needed to ship the key scenarios as a true MVP and surface funnel signals immediately, not after a six-month build.
Non-technical stakeholders
The school's offline operating model had to be translated into precise product rules — what counts as activation, how schedules and seat limits work, how missed sessions and reschedules are handled.
Live operations as a high-risk surface
Time zones, seat limits, reminders, attendance statuses, and session links all had to work reliably so the live experience would feel smooth rather than frustrating.
Mentoring as a product, not a private arrangement
Profiles, specializations, availability, booking rules, and interaction history are what let mentoring scale and hold its price; without them, mentoring collapses into informal coordination that breaks at any volume.
Proposed Solution
We built a headless architecture where Directus serves as both the operational center and the business-logic core. The system concentrates data, roles, access rights, schedule, bookings, progress, and all process states inside Directus. Node.js handles a thin integration layer for webhooks, notifications, and payments. Next.js delivers the product shell — user dashboards and learning flows.
Directus as the domain core
We modeled the key entities — tracks, modules, live events, mentors, slots, bookings, packages, access rights, progress, statuses. Processes run through explicit states (registered → attended → missed, and created → confirmed → rescheduled → cancelled) so live operations and mentoring could scale predictably as a system rather than as ad-hoc coordination.
Access rights and monetization
To remove manual confirmations, we tied payments directly to access rights in Directus — each package defines its access boundaries (tracks, live-session caps, mentor slots, validity period), and active user entitlements live as separate first-class objects in the system.
Live sessions and mentoring
We implemented live registration and mentor booking as controlled processes in Directus — with seat caps, slot limits, deadlines, time zones, change history, and policies for rescheduling and cancellation. Node.js was set up to send reminders and transactional emails based on Directus status changes.
Slots and bookings
We modeled mentor availability as time slots and to make booking a status-driven process with change history and explicit reschedule/cancel policies. This converts mentoring from a private arrangement into a repeatable, priceable service.
Payments and automatic access
We integrated Stripe so that a successful payment webhook creates or activates access rights in Directus for the purchased package. This removes manual confirmations and shrinks the time between payment and learning start to seconds.
Notifications and reminders
We implemented automated notifications around Directus status changes — booking confirmation, pre-session reminders, schedule changes, reschedule confirmations. Node.js was set up as the delivery dispatcher; trigger rules live in the process states inside Directus.
Event analytics
We identified the events that matter for a learning funnel — track selection, track start, live booking, attendance, mentor booking, checkout, payment — and committed to capturing them from day one. The goal was a measurable funnel that supports surgical step-level improvement, not directional guesswork.
We started by extracting the school’s operating mechanics directly from the co-founders. Together, we broke down what a “track” actually means — how it starts, how a learning week is structured, where practice belongs, where live sessions matter, and when mentor involvement becomes necessary.We also mapped every point where the offline process depended on people manually stepping in: reminders, confirmations, reschedules, and handling edge cases ranging from lateness to lost session links.From there, we aligned on the core MVP objective: guide every user to the first meaningful value action — starting a track, booking the nearest live session, or interacting with a mentor.We then defined the minimum operational rules for the platform: seat allocation per group, booking deadlines, reschedule and cancellation policies, attendance handling, and absence tracking. The outcome was a compact operational rulebook that became the foundation for both the Directus data model and the dashboard workflows.
+4 Resource
Results
After launch, the platform became a measurable channel for both learning and revenue generation. Since the product was built from zero, success was defined by how quickly the funnel, operational workflows, and initial monetization stabilized on real traffic.
2.8–4.6%
conversion from organic + social traffic into registration
35–48%
of registered users reached their first value action — track start or live-session booking — inside the product
6–10%
of registered users converted to their first paid purchase
−10–15 hours/week
of manual founder operations removed by automated reminders, status-driven flows, and Stripe-triggered access activation
1–2 product + marketing iterations per week
sustainable cadence enabled by the event-analytics layer and the in-CMS rule control
Live sessions stopped being chat-coordinated
became a repeatable system with predictable seat management, time-zone handling, and automatic notifications
Mentoring became a priceable, scalable service
slot-based booking with reschedule/cancel policies replaced the previous private-arrangement model
Get in Touch
Or send us a message and we'll get back to you within 15 minutes during business hours