Back to ProjectsAI-Powered Educational Platform with HLS Streaming

    Lancémonos

    Video course platform with automatic AI transcription, semantic topic analysis, intelligent search (Algolia), and HLS streaming that solves memory issues with long videos. 276+ videos processed in production.

    276+Videos processed
    < 200msAlgolia search
    -90%Indexing time
    Lancémonos Nomás

    Overview

    How the system works

    Lancémonos Nomás transforms hundreds of hours of educational video into indexed, searchable, and accessible content. An AWS Step Functions pipeline processes each video: extracts audio, transcribes in parallel with AI, analyzes topics and keywords with specialized AI models, generates formatted content, and converts to HLS for efficient streaming. Algolia syncs transcriptions and topics for instant search in Spanish. HLS streaming solved a critical issue: 30+ minute videos caused memory crashes on 4GB devices (500MB+ progressive video vs 36-85MB with HLS). Admin panel for content management with drag-drop and atomic deletion.

    Web PlatformAWSStep FunctionsAI TranscriptionAI AnalysisAlgoliaHLS StreamingDynamoDBCloudFront
    The Challenge

    The Challenge

    276+ course videos unindexed and unsearchable. Long videos (45-90 min) caused memory-related crashes on budget devices. Teachers received repetitive queries that existing content already answered.

    The Solution

    The Solution

    Step Functions pipeline: download video → extract audio → parallel AI transcription in chunks → semantic topic and keyword analysis with AI → HLS conversion with codec detection → HTML generation → Algolia sync. HLS streaming keeps memory stable on any device. Admin manages content with drag-drop reordering.

    Results

    Results

    276+ videos processed in production (556GB). Average search < 200ms with Algolia. HLS streaming eliminated crashes on low-memory devices. Engagement +300%. Costs optimized: sa-east-1 → us-east-1 migration saved 58% ($126/year). S3 lifecycle rules reduce storage costs 65-71%.

    Screenshots

    Discover the project

    Technical Architecture

    Technology Stack

    Next.js web platform
    AWS CDK for complete infrastructure
    Step Functions for processing pipeline
    Lambda: 12 specialized functions
    AI transcription for Spanish (parallel chunks)
    AI-powered topic analysis, keywords, and HTML generation
    Audio extraction and HLS conversion
    Algolia for full-text instant search
    DynamoDB + S3 (556GB, lifecycle rules)
    CloudFront CDN with 365-day cache for HLS
    SQS + SNS for async processing and notifications
    Adaptive streaming on frontend

    Key Benefits

    Improvements after implementation

    Videos processed

    Videos processed

    276+ — Videos processed

    Algolia search

    Algolia search

    < 200ms — Algolia search

    Indexing time

    Indexing time

    Student retention +300% with same teaching resources. 40+ hours/month saved on repetitive teacher queries. AWS cost optimized to ~$7-8/month after region migration.

    Quote Icon
    CloudForgeAI transformed our educational platform by implementing an AI system that automatically processes our videos. The intelligent search answers queries citing exact videos where the required content is shown, elevating the learning experience and increasing student retention by 300%.
    Leonardo TobarCEO, Lancemonos Nomás
    Case Studies

    Discover more case studies

    DexaVision

    AI-powered dental diagnosis in under 30 seconds

    Dental Caries Detection with AI

    AI dental caries detection platform that analyzes images using ICDAS II classification, connects patients with verified dentists via real-time chat, and processes subscription payments. All from a mobile device.

    < 30s

    Full AI analysis

    4 models

    Multi-model pipeline

    45+

    Lambda functions

    View full caseexternal

    Sugeter

    Plan, share and discover travel itineraries in minutes

    Social Travel App

    Instagram-style social platform for travelers where users can create itineraries, add activities, locate points on a map, and share experiences in a guided, fluid, and frictionless flow.

    70%

    Time reduction

    100%

    i18n Ready

    0→1

    Scalable architecture

    View landing pageexternal