A production-grade, fully branded education platform, live payments, in-browser creative tools, microservices architecture, and zero lost enrolments. Here's exactly how we built it, and you can try the Starter Edition free today.
The client needed a platform no SaaS product could offer, complete ownership, custom branding, and a payment system they could trust with real revenue.
The Challenge
A growing education company had outgrown every generic LMS on the market. Teachable, Thinkific, Moodle, none of them could support a vision where teachers record professional lessons, build interactive mind maps, publish content to social media, conduct live whiteboard sessions, and sell courses and tool subscriptions, all under one roof, fully owned, fully branded.
They came to us with one clear requirement: build something that scales, doesn't break when payments happen, and is secure enough for real money. No compromises. No vendor lock-in. No half-measures.
What We Built
A production-grade, multi-service education platform serving teachers, students, and administrators with live payments, automated lifecycle emails, fuzzy course search, personalised recommendations, and a full in-browser creative suite.
Who It's Built For
Corporate L&D teams and higher-ed institutions are the primary fit, both run structured, recurring training or coursework at a scale where "just use a shared drive and Zoom" stops working, and both need content they own outright, not rented seats in someone else's SaaS. Schools run on the same platform too. Off-the-shelf tools force a choice between owning your content and moving fast; with AI-assisted course generation now live, you get both, a full course ships in minutes, not weeks.
Why Teams Switch
Most learning management systems were designed to host courses, not to help people make them. These are the problems training teams describe most often, and how this platform handles each one.
A larger organisation pays several times more for the same platform, and the bill rises with each new employee.
How we solve it: The Enterprise Edition is sized for your teams and learners, with no per-seat surprises as you add departments.
Video lives in one product, written lessons in another, and quizzes in a third, so every course is a juggling act.
How we solve it: Recording, editing, whiteboard, mind maps, quizzes and publishing all run in one browser tab.
You can see that learners stop, but not where they lose interest or which lesson needs work.
How we solve it: Progress is tracked lesson by lesson, with completion and quiz scores per learner, so weak spots are easy to find.
A five-minute refresher still becomes a thirty-minute module because building anything shorter takes the same effort.
How we solve it: AI course generation turns a topic or your existing notes into a structured course as PDF handouts, PowerPoint or interactive slides.
Anything beyond the vendor template needs a paid engagement, so the platform shapes your training instead of the other way round.
How we solve it: Run the platform on your own domain under your brand, or have us build a version tailored to your industry.
Teams either hire a studio or ask trainers to become video editors on top of their day job.
How we solve it: The Recording Studio handles screen and camera capture, cuts, subtitles, noise removal and music, with a teleprompter for the script.
System Architecture
Every service owns its domain, its database, and its deployment lifecycle. Cross-service communication is fully asynchronous via Apache ActiveMQ, a slow recommendation engine can never cause a payment to fail.
Full system architecture, React SPA · Spring Cloud Gateway · 7 microservices · Apache ActiveMQ · pgvector · AWS S3 / SES
Microservices
| Service | Responsibility |
|---|---|
| API Gateway | JWT validation, rate limiting (20 RPS / 40 burst), CORS, routing to all downstream services |
| User Service | Registration, login, JWT minting, password reset, email verification |
| Course Service | Course CRUD, chapters / modules / resources, ratings, pg_trgm fuzzy search |
| Learning Platform | Enrolment, per-lesson progress tracking, completion signals via ActiveMQ |
| Commerce Service | Cart, orders, Razorpay payments, tool subscriptions, coupons, HMAC verify |
| Media Service | File upload to S3, transactional email via AWS SES, lifecycle notifications |
| Recommendation Service | Personalised course recommendations, "Continue Learning" feed, vector similarity |
Engineering Depth
This isn't a naming convention, it's a hard boundary between business logic and infrastructure that paid off when we swapped the mock payment gateway for live Razorpay: zero domain code changed. One adapter file.
Enforce all business rules. ToolSubscription is a state machine, PAYMENT_PENDING → ACTIVE → EXPIRED / CANCELLED / PAYMENT_FAILED. No invalid state is representable in code.
Defined in the application layer, implemented in infrastructure. The domain has zero dependency on JPA, Spring, or any framework. Swap PostgreSQL for MongoDB, domain stays untouched.
Orchestrate use cases. They never reach into persistence or HTTP directly. BookAppointmentUseCase, EnrolStudentUseCase, each owns one workflow, nothing more.
Handle JPA entities, S3 uploads, JMS publishing, and Razorpay calls, fully replaceable without touching business logic. Hexagonal architecture enforced from day one.
Payments & Reliability
The most important architectural decision in the entire project. When a payment succeeds, two things must happen atomically: save the payment and trigger enrolment. If the message broker goes down between those two steps, the payment is recorded but the student is never enrolled. We solved this completely.
Razorpay confirms. Commerce service writes to outbox atomically with payment record.
Scheduled poller reads undelivered events from the outbox table.
Event forwarded to ActiveMQ. Marked delivered only on successful publish.
Learning platform consumes event and auto-enrolls the student.
If the broker goes down, the outbox accumulates. When it comes back, every event is delivered in order, exactly as written. Zero ghost enrolments since launch.
Every event listener is idempotent. A student receiving the same CoursePurchasedEvent twice is enrolled exactly once, no duplicates, no side effects.
Payments
Card data never touches our servers. Every payment callback is cryptographically verified server-side before any access is granted. Browser manipulation cannot bypass subscription checks.
POST /checkout, Commerce Service creates an Order + Payment record, calls Razorpay Orders API, returns razorpayKeyId + gatewayOrderId.
Razorpay Checkout modal opens in the browser. No card data ever touches our servers, PCI scope eliminated entirely.
Frontend calls POST /payments/verify. Backend computes HMAC-SHA256(keySecret, orderId|paymentId).
If signature matches, payment marked confirmed and outbox event fires. If it doesn't match, access is denied, regardless of what the frontend claims.
In-Browser Creative Suite
Everything a teacher needs to record, explain and publish a lesson, purpose-built for education and running in the browser. Nothing to install.
Record screen and camera, then edit in the same place: cuts, titles, images, picture-in-picture, automatic subtitles, background music, noise removal, colour, and splitting one recording into several lessons.
Type the steps and the flowchart draws itself, with decisions, branches and loops. Then present it one step at a time.
Turn a flowchart into a rotating 3D diagram, or explain with ready-made 3D models, and record it straight into a lesson.
Annotate over uploaded PDFs and images, talk through it while you draw, and export the board as an image or course resource.
Keyboard-first mapping (Tab, Enter, F2) with auto-save, and export to an image for slides and handouts.
Draw shapes, arrows and text on top of a screenshot or photo to make quick explainer visuals.
Read your script while you record, scrolling at your pace, so every take sounds prepared.
Write once, adapt for LinkedIn, Facebook, Instagram, X and Threads, and publish everywhere in one click.
EnterprisePlus everyday utilities: trim and crop video, capture thumbnails, resize images, merge and split PDFs.
An optional free desktop app speeds up exports of long videos.
Editions
Try the tools yourself today, run the full platform under your own brand, or have us build something like it for your business.
For teachers and creators
Sign up and start creating today. No setup, no installation, free to start.
For companies, colleges and schools
The full platform, on your own domain and under your brand, sized for your teams and learners.
For businesses with a product idea
Need a platform like this for your industry: training, onboarding, a marketplace, a SaaS tool? We design and build it with the same production-grade engineering.
Live in Starter · Workflow Tool
Teachers describe a process as a plain outline, one step per line, and the Workflow tool turns it into a real flowchart, with decisions, branches, loops, databases and API calls laid out automatically. Then it presents the chart one step at a time, ready to record or drop into a lesson. No AI and no drawing: it all runs in the browser.
Work in progress: everything shown in this demo works today, but the interface is still being refined. The toolbar is being reorganised into a cleaner, less crowded layout, so the finished tool will look tidier than this recording.
The value: end a line with ? for a decision, indent no: Fix it for a branch, add -> Step to loop back. The chart is laid out instantly, and it can be edited by dragging or as text, whichever is quicker.
The value: start and end, steps, yes/no decisions, input/output, databases, API calls, users and waits, each drawn as its standard flowchart symbol. Technical processes read correctly at a glance, not as a row of identical boxes.
The value: present mode builds the chart the way a teacher would explain it, one step per click, with arrows drawing themselves in and a choice of pop, fade, slide, zoom or swirl. Learners follow the logic instead of facing the whole diagram at once.
The value: one click records the build and the presentation straight into the Recording Studio for editing. Export a 1920×1080 image for every step, turn the chart into a rotating 3D diagram in 3D Studio, or add it to an interactive slide where learners click through it themselves.
Record the way you normally teach. Auto listens to what you explain and adds what helps people follow you, right where you say it. You review and edit every suggestion before it goes into your video.
Get Early AccessStarter users will see it in the Recording Studio first.
Latest Update · AI Course Generation
A teacher types a topic, or pastes existing notes, and the platform's AI planning pipeline designs a complete, pedagogically structured outline, objective, theory, worked example, exercise, assessment, then generates real content for every slide. No blank page. No generic template.
Type a subject, or paste what you already teach from.
A structured plan, objective through assessment, not a wall of bullet points.
Every slide, example, and quiz question, written to that plan.
The same course, delivered the way it'll actually be used.
Clean, printable material, headings, worked examples, and code, formatted like a proper document for reading offline or handing out before a session.
Why it matters: takeaways students keep, zero design workA real .pptx file with the same content, structured slide by slide for live teaching. Download it, edit it in PowerPoint, present it anywhere.
Why it matters: walks into any classroom or conference room as-isA genuine in-browser presentation, not a PDF wearing a slide deck's clothes, diagrams that animate into place, flowcharts learners reveal one step at a time, motion on every image, and scored quizzes built in.
Why it matters: students engage and get graded, not just scrollWhat sets this apart: diagrams and workflows assemble on screen node by node, exactly like a presenter building them live, not a static image. Images get real motion, Ken Burns pans, tile-scatter reassembly, focus pulls, chosen automatically or picked by the teacher.
What sets this apart: every interactive deck includes at least one real, scored knowledge check, generated from what that specific course actually taught, never generic trivia, with per-question explanations and a running score as students answer.
You're Always in Control
The AI-generated outline and content are a starting point, not a locked box. Every chapter, slide, and word stays fully editable, and you can redirect the AI itself instead of starting over from a blank prompt.
The value: the AI-generated outline is a first draft of the course structure, not the final word. Add a chapter it missed, merge two modules, delete a slide you don't need, reorder anything. Your course, your structure.
The value: type an instruction, "make this more technical," "add a worked example here," "simplify for beginners," and regenerate a single slide or the whole lesson to match. No re-prompting from scratch.
The value: every heading, paragraph, quiz question, diagram, and image is directly editable in the browser. No AI round-trip needed for a quick fix, and no risk of an AI edit undoing work you did manually.
The value: theme, background texture, per-image motion style, and quiz behaviour are all teacher-controlled, per slide or across the whole deck. AI picks sensible defaults; you get the final say.
See It In Action
Live screens from the running platform - the marketing site, the course catalog, and the in-browser studio tools, opened as windows on top of the app.
The marketing entry point - public or private publishing, built-in creative tools, and a live course preview.
Search and category filtering over published courses, backed by pg_trgm fuzzy search.
A full tldraw-based canvas for live teaching sessions, opened as an overlay without leaving the page.
Record screen and camera, then edit in the same window: titles with Word Art, cuts, picture-in-picture, subtitles, music and a timeline.
Type the steps and the flowchart lays itself out, with decisions, branches and loops, numbered in the order they're revealed.
The same flowchart as a 3D diagram you can tilt, rotate and record, or download as a 3D model.
Compose once, adapt with AI, and publish across LinkedIn, Facebook, Instagram, X, and Threads.
Text-to-speech across languages and genders, plus custom voice cloning for narrated lessons.
AI Recommendations
The recommendation service uses a hybrid approach, vector cosine similarity via pgvector as the primary engine, with pg_trgm similarity as a fallback for keyword and tag matching. Both layers combine with full-text search and business filters.
cosine_ops via pgvector
Results
Every architectural decision was made to protect revenue, reduce operational overhead, and deliver a platform the client's team can own and extend independently.
Draft → Teacher submission → Admin approval → Published, with rejection reasons and resubmission flow built in.
Server-side access checks on every tool page. Browser manipulation cannot bypass a paywall, only a valid DB row grants access.
Expired subscriptions at 01:00 UTC, reminder emails at 02:00, orphaned payments at 03:00, outbox cleanup at 03:30. Zero manual intervention.
Roadmap
This is an ongoing engagement, the core platform is built and stable. New features are being designed and shipped continuously. Here's where things stand today.
Technology Stack
How We Build
Traditional recommendation systems rely heavily on tags and keyword overlap, this often misses deeper contextual relationships between learning content. Here's how we thought differently.
Tag and keyword-based matching surfaces courses that share labels, not courses that share meaning. A student interested in "neural networks" gets courses tagged "AI" but not the foundational "linear algebra" course they actually need next.
Semantic embeddings via Sentence Transformers + pgvector in PostgreSQL with cosine similarity search. Embeddings capture contextual meaning, not just surface keywords. Combined with event-driven ingestion pipelines for continuous freshness.
We optimised for operational simplicity, scalability, and cost efficiency without prematurely introducing unnecessary infrastructure complexity. pgvector with HNSW index delivers sub-100ms similarity search at our scale, no separate vector database needed.
The platform delivers more relevant recommendations, contextual learning discovery, and improved personalisation, through a scalable pipeline that evolves responsibly with scale. Good architecture isn't about using the most technologies. It's about building systems that grow cleanly.
Architecture Snapshot
cosine_ops
Good architecture is not about using the most technologies. It's about building systems that evolve responsibly with scale.
How We Build
As the platform evolved with browser-based recording, clip editing, timeline stitching and async exports, traditional in-memory processing using ConcurrentHashMap became a hard limit on horizontal scaling. Here's how we redesigned the architecture around Redis-based distributed coordination.
In-memory job state breaks the moment you run more than one server instance. Two workers can simultaneously pick the same job, a server restart loses all in-flight processing, and a client on instance A can't see the state of a job running on instance B.
Distributed coordination via Redis: LPUSH/BRPOP for job queues, atomic SET NX EX for distributed locking, HSET/HGETALL for shared job state, and Pub/Sub channels for real-time progress to clients via WebSocket.
Redis gave us sub-millisecond performance, atomic operations, TTL-based auto-cleanup, and Pub/Sub, all in a single lightweight infrastructure component. No Kafka or ActiveMQ overhead at this scale. Stateless workers mean we add instances with zero configuration change.
Multiple video worker instances run safely across servers without race conditions or duplicate processing. Clients receive live progress updates without polling. Lock TTL auto-expires on crash, jobs re-enter the queue automatically. Horizontally scalable from day one.
Architecture Snapshot
LPUSH / BRPOP video:jobsSET lock: {videoId} NX EX 30HSET job:{id} status progressINCR ratelimit:{userId} EX 60
Build lean initially. Design for scale from day one.
What Sets This Apart
Plenty of tools will generate a pretty slide deck from a prompt. Here's what each part of this one is actually built to do, and why that's worth paying for over a generic AI slide tool bolted onto your LMS.
The value: a pedagogically structured outline comes first, objective, theory, worked example, exercise, assessment, the same instructional design a course architect would use, not a slideshow of AI-generated bullet points. Students learn in the right order, every time.
The value: every interactive deck can include scored quizzes tied to exactly what was taught on the slides before it, with per-question explanations and a running score. No separate quiz tool, no generic question bank.
The value: generate once, get a PDF for handouts, a PowerPoint for live teaching, and an interactive deck for self-paced learning, each built for how it's actually used, not the same file wearing three different hats.
The value: no vendor lock-in, no separate per-seat AI tool subscription stacked on top of your LMS bill. Course generation lives inside the platform you already own, on the same production architecture behind your payments and enrolment.
FAQ
It is a course creation platform and learning management system in one. Teachers record, build and publish courses in the browser, students buy and follow them with lesson-by-lesson progress, and administrators approve courses and manage pricing, coupons and refunds.
Yes. The Starter Edition is free to start at learning.ramsudtechnologies.com. It includes the Recording Studio with full editing, the Workflow tool, 3D Studio, Whiteboard, Mind Map, Image Editor and Teleprompter, and you can publish and share courses.
The Enterprise Edition is sized for your teams and learners, with no per-seat surprises as you grow. Book a demo call to discuss the size and needs of your organisation.
Yes. The Enterprise Edition runs on your own domain under your branding, and your course content and learner data remain yours.
Yes. In the Enterprise Edition, a teacher enters a topic or pastes existing notes, and the platform plans a structured course and generates it as PDF handouts, PowerPoint or interactive slides. Teachers review and edit everything before it is published.
Corporate L&D teams and colleges or universities are the primary fit. Schools use it too, with recording, whiteboard and quiz tools for every subject teacher.
Payments run through Razorpay. Card data never touches our servers, and every payment is verified server-side with an HMAC signature before any access is granted.
Yes. If you need a platform like this for training, onboarding, a marketplace or a SaaS product, we design and build it with the same engineering, fully owned by you. Book a free consultation.
We build custom software for education companies, SaaS startups, and businesses that need more than a template can offer. Tell us what you have in mind, or see our full range of services.
Built for your exact workflow, not a generic LMS with your logo on it.
Tamper-proof, transactionally safe, and verified server-side. No lost transactions.
Designed for independent deployment, owned by your team, built to scale.
Recording studio, workflow, 3D, whiteboard and publisher, integrated into your product.