iVisuality private preview
Verified 2026-09-17

Every skill, working.

Not a roadmap of intentions — the capabilities that actually run today, grouped by the stage of the pipeline they power. LIVE = shipping in production now.

123
skills across the stack
455
explainers built with them
1,654
narration clips produced
4
subjects proven end-to-end

Subject library

Know the topic, its curriculum place, and its vocabulary.

subject_packets.pyREADY
8-subject taxonomy spine, 10-slot packets
NAMING_ENUM.pyREADY
closed 12-tab content grammar + audio filenames
catalogue_data.pyLIVE
curriculum-mapped experiment catalogue (host_a)
CAE exam catalogueREADY
~70 exams, topic hierarchy
unified subject indexIN BUILD
one queryable spine across sources

Pedagogy

Explain a complex topic clearly, correctly, at the right level.

pedagogy_pipeline.pyLIVE
10-subject LLM cleaner + intuition + real-world examples
lab-pedagogyLIVE
governing rule: research the real thing, then teach
chem-lab-create-experimentLIVE
topic → structured ordered spec
golden_gate.pyLIVE
ship gate > 0.97
Nova plan stepIN BUILD
swap Haiku → Nova for the plan

Visualisation

Draw it as a real, interactive, animated 2D/3D scene — not clip-art.

chem/phys/bio/math lab enginesLIVE
deterministic shaded 3D solids
reaction_server.pyLIVE
live 3D render service
3d-bio-visualiseLIVE
medically-accurate parametric anatomy
procedural-creature-glbLIVE
self-made GLB organs/creatures
model-svg-showdownREADY
multi-model SVG generation bench
flatness gateLIVE
auto-reject flat/fake 3D
OSS renderersTO ADOPT
three.js/R3F, anime.js, Plotly, KaTeX, Mermaid, 3Dmol, Rapier

Speech

A voice that explains well and pronounces every term correctly.

synth_elevenlabs.pyLIVE
EN narration synth
chem-lab-create-voiceLIVE
EN + HI, 1:1 with steps
bhashini-long-audio chunkingLIVE
beat the char ceiling, one clean MP3
detecting-truncated-ttsLIVE
narration QA across languages
voice provider routingREADY
Bhashini/Sarvam/ElevenLabs, 14 langs
pronunciation cleanerIN BUILD
extract as standalone stage

Sync + delivery

Lock audio to motion, then ship it in any format.

per-step MP3 alignmentLIVE
1,654 clips aligned to steps
player-controls gateLIVE
verified scrub/play sync
audio-player.jsREADY
drop-in narration UI
entitlement.mint_linkREADY
signed, TTL-bound asset URLs
multi-format renderersROADMAP
embed / MP4+VTT / GLB+USDZ / SCORM / xAPI / LTI / source-JSON

Missing something?

Tell us what capability you need. We prioritize by demand.