About Us
Anelo
About Us
Anelo
modular media toolchain built to turn 2D into 3D/VR.
We exist to remove friction in 3D/VR supply. THIS IS WHY WE GIVE AWAY BASIC EDITING, UPSCALING, AND INTERPOLATION.
Anelo re-stimulates 2D media using upscaling, interpolation, depth + stereo, AND stabilization to ACCESS UNTAPPED value in MEDIA you already have.
We seek to enable creators, studios, advertisers, educators, social media companies, and editors WITH tools to be first-movers in CIRCULAR media.
We’re opinionated about one thing: quality beats hype.
That’s why Anelo is built around preview-first iteration, repeatable recipes, and privacy-safe exports — creators, studios, and teams can SCALE PRODUCTION RELIABLY to earn more and reach more from pre-existing assets.
modular media toolchain built to turn 2D into 3D/VR.
We exist to remove friction in 3D/VR supply. THIS IS WHY WE GIVE AWAY BASIC EDITING, UPSCALING, AND INTERPOLATION.
Anelo re-stimulates 2D media using upscaling, interpolation, depth + stereo, AND stabilization to ACCESS UNTAPPED value in MEDIA you already have.
We seek to enable creators, studios, advertisers, educators, social media companies, and editors WITH tools to be first-movers in CIRCULAR media.
We’re opinionated about one thing: quality beats hype.
That’s why Anelo is built around preview-first iteration, repeatable recipes, and privacy-safe exports — creators, studios, and teams can SCALE PRODUCTION RELIABLY to earn more and reach more from pre-existing assets.
Quality Over Hype
We’re building for outputs that hold up in motion — not demo reels. That means preview-first workflows, refinement steps, and QA/QC that reduce artifacts before you export.
Control Without Complexity
Power tools shouldn’t feel like a cockpit. Anelo stays modular and configurable, but the defaults, recipes, and UI are designed to make “good results” the easy path.
Privacy by Default
Media carries more than pixels. We surface and remove sensitive metadata (including GPS) before export so you can share confidently — without accidental oversharing.
Our Journey
The Anelo Adventure
Q1 2025 (Feb–Mar)
Spark
Started from the practical question: why can’t a normal person watch a hit show like Bluey in 3D/VR? Validated the obvious baseline: the core model classes already exist (upscale, interpolation, depth, stereo), but making them behave as a coherent workflow is the hard part. Built initial “recipe experiments” across combinations and ordering (upscale vs interpolate vs depth vs stereo), and hit the core constraint fast: model interoperability is basically not a thing — outputs don’t standardize, assumptions differ, and quality failure modes compound. Established early operating rules: preview early, iterate fast, and treat “good on a frame” as irrelevant unless it holds up in motion.
Q2 2025 (Apr–Jun)
Toolchain direction: recipes over one-off pipelines
Shifted from ad-hoc scripts to a modular stage / toolchain approach so workflows could be composed, swapped, and improved without rewriting everything. Began formalizing repeatability: consistent inputs / outputs, basic run records, and the beginnings of “manifests” (what ran, with what settings, producing which artifacts). Focused on the core UX reality: users don’t want 30 knobs — they want outcomes (comfort-first vs depthier vs faster) and clear tradeoffs.
Q3 2025 (Jul–Sep)
Making it usable: desktop-first iteration loop
Committed to desktop-first as the fastest path to a usable MVP: local control, quick previews, and tight iteration for messy real-world media. Hardened the workflow loop: input → preview → tune → export with fewer dead ends and clearer “what to do next” paths. Continued expanding stage coverage and wiring: more consistent handoffs between enhancement, temporal steps, depth, and stereo— less “magic,” more predictable plumbing.
Q4 2025 (Oct–Dec)
Reliability push: don’t waste compute
Invested heavily in correctness and output integrity: fewer silent failures, fewer “it finished but it’s unusable” outcomes. Strengthened the “don’t waste compute” principle with early versions of preflight checks and quality gating (reject / route before heavy spend). Improved quality tooling where it actually matters for comfort: stabilization / temporal consistency, and edge handling became standard focus areas, not afterthoughts.
Q1 2026 (Jan–Present)
Hardening & MVP
Tightened validation and QA / QC logic so recipes behave more deterministically and failures are explainable (not mysterious). Elevated privacy posture: clearer metadata visibility and sensitive-data redaction (including GPS) as default-safe behavior. Prepared early access posture: MVP that runs locally with hybrid execution (desktop-first, cloud when needed / wanted).
Our Journey
The Anelo Adventure
Q1 2025 (Feb–Mar)
Spark
Started from the practical question: why can’t a normal person watch a hit show like Bluey in 3D/VR? Validated the obvious baseline: the core model classes already exist (upscale, interpolation, depth, stereo), but making them behave as a coherent workflow is the hard part. Built initial “recipe experiments” across combinations and ordering (upscale vs interpolate vs depth vs stereo), and hit the core constraint fast: model interoperability is basically not a thing — outputs don’t standardize, assumptions differ, and quality failure modes compound. Established early operating rules: preview early, iterate fast, and treat “good on a frame” as irrelevant unless it holds up in motion.
Q2 2025 (Apr–Jun)
Toolchain direction: recipes over one-off pipelines
Shifted from ad-hoc scripts to a modular stage / toolchain approach so workflows could be composed, swapped, and improved without rewriting everything. Began formalizing repeatability: consistent inputs / outputs, basic run records, and the beginnings of “manifests” (what ran, with what settings, producing which artifacts). Focused on the core UX reality: users don’t want 30 knobs — they want outcomes (comfort-first vs depthier vs faster) and clear tradeoffs.
Q3 2025 (Jul–Sep)
Making it usable: desktop-first iteration loop
Committed to desktop-first as the fastest path to a usable MVP: local control, quick previews, and tight iteration for messy real-world media. Hardened the workflow loop: input → preview → tune → export with fewer dead ends and clearer “what to do next” paths. Continued expanding stage coverage and wiring: more consistent handoffs between enhancement, temporal steps, depth, and stereo— less “magic,” more predictable plumbing.
Q4 2025 (Oct–Dec)
Reliability push: don’t waste compute
Invested heavily in correctness and output integrity: fewer silent failures, fewer “it finished but it’s unusable” outcomes. Strengthened the “don’t waste compute” principle with early versions of preflight checks and quality gating (reject / route before heavy spend). Improved quality tooling where it actually matters for comfort: stabilization / temporal consistency, and edge handling became standard focus areas, not afterthoughts.
Q1 2026 (Jan–Present)
Hardening & MVP
Tightened validation and QA / QC logic so recipes behave more deterministically and failures are explainable (not mysterious). Elevated privacy posture: clearer metadata visibility and sensitive-data redaction (including GPS) as default-safe behavior. Prepared early access posture: MVP that runs locally with hybrid execution (desktop-first, cloud when needed / wanted).
