Labs→CodeAiiQ
Developer Education · Video ProductionLive

CodeAiiQ — coding tutorials, authored as code.

A visual IDE and deterministic video compiler for coding-tutorial production. Write the tutorial once as a structured script; render it forever — reproducibly, identically, on any platform. Pronounced Code-ek.

AudienceDevRel, developer educators, technical content teams
PlatformsWindows · macOS · Linux
EngagementLicense · Pilot · Contact us
The Problem

Coding tutorials are expensive to make and nearly impossible to update.

A single ten-minute coding tutorial swallows hours of screen recording, dozens of re-takes when a command is fumbled, careful audio post-production, and a full re-record every time the underlying code, library version, or script changes. General video tools — Camtasia, Premiere, After Effects — were built for film-making. They do not understand terminals, editors, or the pace of a coding narrative.

CodeAiiQ flips the model. Instead of recording a performance, you write the tutorial the way engineers write software: scenes, shots, terminal commands, editor operations, and narration — all declared explicitly in a version-controlled script. A deterministic compiler turns that script into a polished .mp4. When the code changes, you update the script and re-render. When the narration changes, you swap the voice and re-render. The tutorial is a build artefact, not a one-time take.

How it works

Script. Compile. Render. Ship.

01

Author

Open the visual IDE. Lay out scenes and shots, type narration, drop in terminal commands, editor operations, overlays, and clips. WYSIWYG authoring with a live diagnostics engine catching issues as you type.

02

Compile

Save and the project is committed to a MessagePack .aiiq file. A deterministic export emits a v3 Markdown contract — the single source of truth the engine renders from. Script-as-source means every change is diffable and version-controllable.

03

Render

The Studio Engine runs a headless Chromium that renders the tutorial inside a real IDE-like environment. TTS narration (Edge TTS, ElevenLabs, or OpenAI) is synthesised and synced frame-accurate to visual actions. Each scene records in isolation.

04

Ship

FFmpeg assembles the scenes, normalises audio, applies transitions, and outputs a polished .mp4. Same input next week? Same output. Swap a voice, tweak a code snippet, re-run — no re-recording, no re-editing.

What makes it different

Four decisions baked in from day one

◆

Deterministic

Identical input always produces identical output. Re-rendering last month's tutorial yields the exact same frames — perfect for series consistency, localisation, and versioning.

◆

Explicit, not inferred

Every visual action is declared in the script. No hidden magic, no flaky playback. Twenty-one action types cover terminals, editors, browsers, overlays, stages, flows, and media.

◆

Pipeline isolation

Parse, TTS, render, record, and assemble run as independent stages. Re-run just the stage you need — regenerate narration without re-rendering, tweak assembly without re-running TTS.

◆

Secure by design

Sandboxed Electron renderer, context-isolated preload, and OS-keychain storage for paid-provider API keys. Plaintext never leaves the main process; TTS subprocesses receive keys through env, never argv.

Philosophy

The machine was always there.

CodeAiiQ is a deliberate act of ownership. We already have the machines. We already have the open source. We already have the intelligence. The only thing we have been conditioned to lack is the belief that we can build tools on our own terms.

Open source to freedom. Every core part of CodeAiiQ runs on battle-tested open source — the kind millions of engineers already trust. No vendor lock. No proprietary black boxes. No API we cannot replace next quarter. You own what you ship; so do we.

Only IQ. No AI. CodeAiiQ is not an LLM wrapper dressed up with a UI. The intelligence here is engineered — deterministic state machines, explicit action taxonomies, a reducer-driven visual renderer. No hallucinations to second-guess. No inference cost per tutorial. No model provider we depend on to keep the lights on. AI built the tool. Then it stepped aside.

Proven parts, unproven applications. Playwright is the world's favourite framework — for testing web apps. We drive it instead to orchestrate a real IDE-like environment that records itself frame-perfect, replacing an entire category of AI video generators. The parts are ordinary. The composition is the product.

We bought the land. We refuse to rent the garden. Read the longer case at iq.aiiqlabs.com. CodeAiiQ is what that thesis looks like when you ship it.

Status & Roadmap

Shipped today. Shipping next.

Live today

Full visual IDE with 3-panel scene authoring, timeline zoom, and multi-select batch operations. Deterministic compilation pipeline with 21 action types. Multi-provider TTS (Edge TTS, ElevenLabs, OpenAI) with OS-keychain credential storage. Validation engine with 11 rules surfaced through inline cards, status pills, and a jump-to-card flash. Ships cross-platform on Windows, macOS, and Linux.

Shipping next

Production-recording feedback loop — the next cycle of enhancements is driven by real use during active tutorial production. UX refinements, expanded action coverage for common authoring patterns, and performance tuning for long-form scripts.

On the roadmap

Extended scene transitions, richer asset management, additional TTS provider integrations, and first-class collaboration primitives for teams producing tutorial libraries at scale.

Ready to see CodeAiiQ in action?

CodeAiiQ is shipping today on Windows, macOS, and Linux. Book a walkthrough with the team that built it, or talk to us about licensing, enterprise deployments, and tailored integrations for your tutorial production pipeline.