PROJECT

AI video generator for automated content production

Custom software development, Text-to-video generation, Script and media import, Video editing, AI voiceovers and captions, Video rendering, YouTube and social publishing

  • Industry: Content production
  • Lithuania Location: Lithuania
  • Delivery time: 16 weeks
  • Technologies: Python, JavaScript / TypeScript (Node.js)
4+ months

This client is a content agency from Lithuania. Producing short-form videos from scripts, prompts, and articles takes manual work. This includes editing, captions, and publishing. The AI video generator turns these inputs into editable scenes, finished videos, and platform-ready exports. It gives the agency one place to create branded content, prepare metadata, and schedule videos for YouTube and social channels, with the purpose of reducing production time and manual work.

Summary

Problem

The agency spends time moving from scripts and raw media through editing, branding, rendering, and publishing.

Solution

The solution became one AI video creation workspace for source content, scenes, templates, and publishing.

Result

The software is made to reduce manual production work while keeping repeatable video creation in one place.

Client story

The client is a content agency from Lithuania. It produces short-form video variations and branded content for social channels.

Its work includes scripts, prompts, and articles. It also uses images, data, and existing media.

The agency spends time assembling scenes, choosing visuals, and preparing voiceovers. It also edits captions, applies brand settings, and prepares publishing details.

The request was simple: generate videos from varied content sources, keep branding consistent, support repeatable batch production, and publish finished videos through platform connections.

The solution was shaped as a modular content production workspace. Separate screens cover creation, editing, export settings, and the publishing queue.

Documentation helps the agency operate the system independently.

The client asked for

  1. Keep logos, colours, fonts, themes, styles, and templates configurable
  2. Schedule and publish videos to YouTube and social platforms
  3. Support repeatable batch production with minimal manual editing
  4. Generate videos from scripts, prompts, articles, images, data, or media

Why ready-made software did not fit

Ready-made AI video generators cover text-to-video, editing, and distribution, but their descriptions also point to limited credits or high costs. The required system also supports batch production, documented operation, and the required output formats and platforms.

Project overview

The solution was designed around the source content, editing, rendering, and publishing stages described for the agency. The creation workspace accepts scripts, prompts, and topics.

It also accepts articles, data, and images. It accepts audio, video clips, and existing media.

To handle different production styles, the workspace assembles scenes and visuals into an editable timeline. remotion (a library for creating videos programmatically) supports video creation, while FFmpeg (a tool for processing and rendering media) handles media operations.

Special attention was paid to repeatable branding. The system applies logos, colours, and fonts to video outputs.

Themes, styles, and reusable editing templates are also supported.

The publishing queue brings rendered videos together with titles, descriptions, and tags. python-youtube (a Python wrapper for the YouTube Data API) supports YouTube publishing workflows. Automation tools coordinate rendering, scheduling, and publishing.

The studio structured the system as modules so later features, forms, automations, or integrations can be added without rebuilding the whole workspace. Documentation and training support the agency team’s day-to-day operation.

Key features delivered

  • Generate videos from scripts, prompts, topics, articles, images, data, or existing media
  • Assemble scenes, visuals, footage, voiceovers, music, captions, and transitions
  • Edit videos through a scene list, preview, and timeline
  • Apply logos, colours, fonts, themes, styles, and reusable templates
  • Render 16:9, 9:16, and 1:1 videos in selected resolutions
  • Create subtitles, dynamic captions, thumbnails, and publishing metadata
  • Prepare short clips from long videos or video URLs
  • Schedule and publish videos or clips through platform APIs and automation workflows

Can be added

+ AI-generated visuals and logos + Short-clip creation from long videos + Google Sheets content input + Additional social platform connections

Who else can use this solution

Other content-focused businesses have a similar task, and the same solution can be adapted to their production and publishing process.

A YouTube channel operator

Task They turn scripts, topics, and existing media into educational, entertainment, technology, news, sports, or faceless videos.
How the solution helps The solution organizes creation, editing, captions, branding, rendering, and publishing in one workspace.
Creative studios

A marketing and service business

Task It creates branded promotional or product videos for social media.
How the solution helps The solution applies reusable brand settings and prepares platform-ready videos with publishing metadata.
Creative agencies

A publisher

Task It produces large volumes of short-form video variations from structured content and source media.
How the solution helps The solution supports repeatable production, multiple formats, captions, and scheduled distribution.
Creative studios

Technology stack

Python

Content and API services

Python supports source handling, transcript access, and YouTube API services.

JavaScript / TypeScript (Node.js)

Video and automation services

Node.js supports programmatic video creation, media processing, captions, and integrations.

PHP

Media and platform services

PHP supports media embedding and compatible video service connections.

remotion

Programmatic video creation

remotion creates videos from structured scenes and project data.

openai

OpenAI API access

openai connects the software to the OpenAI API for AI-assisted content generation.

ffmpeg

Media processing and rendering

FFmpeg processes video and audio and renders the selected output.

youtube-transcript-api

YouTube subtitle retrieval

youtube-transcript-api retrieves transcripts and subtitles from YouTube videos.

python-youtube

YouTube publishing access

python-youtube connects publishing workflows to the YouTube Data API.

subtitle

Subtitle parsing and editing

subtitle parses and manipulates WebVTT and SRT caption data.

MP4, WebVTT, SRT

Video and caption outputs

The project produces rendered videos and subtitle files for editing and publishing workflows.

Integrations

YouTube Data API

The connection sends video metadata and rendered videos to YouTube and records upload status.

YouTube

The software retrieves transcripts and subtitles from video URLs for short-clip and content workflows.

Google Sheets

Structured content from Google Sheets can enter the production pipeline.

n8n or Make.com

Automation workflows coordinate generation, rendering, scheduling, and publishing.

The result

Start videos from existing content

The first screen gives the agency one place to start a project from scripts, prompts, and topics. It also accepts articles, data, and media for video creation.

The content creation workspace records source text, uploaded media, and project status.

How it works

  1. The input editor holds scripts, prompts, topics, articles, and data.
  2. The media area stores uploaded images, audio, and video clips.
  3. The project list shows the source content and project status.
Content creation workspace

Content creation workspace

Functions on this screen
Create or Import Videos Import Script Generate Video

Shape the story with scenes and captions

The scene and timeline editor brings the assembled video into an editable view. Scenes, visuals, and B-roll appear with the video preview.

Voiceovers, music, subtitles, and captions remain available while the video is edited.

How it works

  1. The scene list holds scene text and generated visuals.
  2. The timeline combines footage, voiceover segments, and music tracks.
  3. The preview and editing panels show subtitles and caption text with the assembled video.
Tip: The editor keeps storytelling elements together before the video is rendered.
Scene and timeline editor

Scene and timeline editor

Functions on this screen
Edit Videos for Storytelling Add voiceovers and subtitles Edit dynamic captions Mix background music

Keep branded outputs consistent

The brand and export screen applies logos, colours, and fonts to the video. Themes, styles, and reusable templates keep the visual treatment consistent.

The format controls support 16:9, 9:16, and 1:1 outputs with selected resolutions.

How it works

  1. The brand controls store the logo, colours, and fonts.
  2. The theme and template controls hold the selected visual style.
  3. The format controls store the aspect ratio and output resolution for rendering.
Brand and export settings

Brand and export settings

Functions on this screen
Video Rendering Automation

Prepare publishing in one queue

The publishing queue keeps the rendered video and its publishing assets together. Titles, descriptions, and tags appear with the video.

Hashtags, thumbnails, and platforms support the final publishing step. The queue also stores schedules and upload status.

How it works

  1. The queue holds the rendered video, thumbnail, target platform, and schedule time.
  2. The metadata fields store the title, description, tags, and hashtags.
  3. The publishing connection sends the video and metadata to the selected platform and records the upload log.
Note: Publishing support covers YouTube and selected social platforms through APIs or automation workflows.
Publishing queue and metadata

Publishing queue and metadata

Functions on this screen
Post videos to Instagram

What the solution gives

With this solution, the agency can:

  • Create videos from scripts, prompts, articles, images, data, and existing media
  • Edit scenes, visuals, voiceovers, music, subtitles, and captions in one workspace
  • Reuse brand settings and templates across different video formats
  • Prepare titles, descriptions, tags, hashtags, and thumbnails with each video
  • Render platform-ready videos and short clips
  • Schedule and publish content through connected platforms and automation workflows

Could ready-made systems do the job?

Before the design, the task was compared with ready-made products by their own descriptions.

What the client needs AI video generators Simplified, AI Studios, Steve AI AI video platforms Hippo Video, Colossyan Creator, Goldcast Video repurposing tools Vizard, Waymark Custom solution
Configurable branding and templates
Schedule and publish videos partly partly
Repeatable batch production partly
Generate from varied content partly
Fits the client's own rules: stages, roles, reports — — —
All steps in one place — —
No manual steps between programs partly partly —

By the public descriptions of the products and their category.

The main parts of the task exist in ready-made AI video products, and one is enough when only text-to-video or basic editing is needed.

This agency needed its own software because the required inputs, branding rules, repeatable production, publishing path, and operating documentation had to work together.

The parts are spread across separate products in some options, and the steps between generation, rendering, storage, scheduling, and publishing may still need automation.

In the end

In the end, the solution was made to take manual work out of the path from source content to a branded, published video.

FAQ

Thinking about building a similar system? Read the answers below to see how a custom AI video generator can fit your content workflow.

Need something similar? Let's talk
Or browse more projects