Press & brand kit

For journalists, bloggers, and analysts.

Logos, approved tagline, executive bio, product screenshots, and the one-line facts sheet. Everything you need to write about LabelSets and the LQS standard without re-describing it to us on a call.

Fact sheet

The facts, in one grid.

  • CompanyLabelsets LLC (Florida, USA)
  • ProductLabelSets — AI training-data marketplace + LQS quality standard
  • Founded2026
  • HeadquartersFlorida, USA
  • CategoryAI training data · data quality · model risk management
  • StandardLQS v3.1 — 19 dimensions · 7-oracle consensus · Ed25519-signed certs
  • Signing keyaa4c070af907e2ea (Ed25519 fingerprint, public)
  • Open sourceScorer + weights ship MIT Q4 2026 · paper Q2 2026
  • Python SDKpip install labelsets · v0.3.0 on PyPI
  • PricingDatasets from $29 · 85% seller payout · enterprise by arrangement
Approved copy

Pre-written descriptions. Copy and paste.

Three length variants so you can drop us into a headline, a paragraph, or a full profile without retyping. All are vetted by us and approved for reprint.

One-liner
LabelSets is the training-data quality standard for AI.
Short — 30 words
LabelSets is a procurement-grade quality standard for AI training data. Every dataset ships with a 19-dimension LQS rating and an Ed25519-signed cert teams can cite in SR 11-7, EU AI Act, and §1557 paperwork.
Long — 90 words (press release intro)
LabelSets (labelsets.ai) is the quality standard for AI training data. The LQS rating — 19 dimensions, scored by a 7-oracle consensus and signed with Ed25519 — turns dataset procurement from a trust exercise into a verifiable one. Enterprises can cite the cert directly in SR 11-7 model-risk paperwork, EU AI Act Article 10 data-governance filings, and FDA 21 CFR 11 audits. The LabelSets marketplace is the proof point: every dataset listed ships with a signed cert. The scorer itself will be released open-source under MIT in Q4 2026.
Logos

Brand marks, in every format.

SVG preferred (scales cleanly); PNG available for editorial pipelines that don't accept SVG. Do not modify the mark or add drop-shadows, strokes, or recolor.

Dark background
LabelSets

Primary logo — dark

Cyan accent bars + white wordmark. Use on dark backgrounds (recommended minimum contrast: #030614).

Light background
LabelSets

Primary logo — light

Same mark rendered with darker fill for print and editorial use on white backgrounds.

Colors
#030614
#22D3EE
#FFFFFF
gradient

Brand palette

Ink #030614 canvas. Cyan #22D3EE single accent — no purple, no green. Gradient reserved for the logo mark and hero headlines.

Typography
Inter — all weights
JetBrains Mono — cert IDs · code · stats
Letter-spacing -0.01 to -0.03 on headings. Tabular numerics on stats.

Type system

Inter everywhere for running copy. JetBrains Mono for anything that looks like a cert, hash, or metric. Both open-source (OFL).

Product screenshots

Hi-res shots for editorial use.

Use the OG card as the default editorial visual — 1200×630, all brand-approved. For specific page captures, grab directly from the live site and retain proportions.

LabelSets homepage OG card
Homepage — labelsets.ai
LabelSets LQS standard OG card
LQS standard — /standard
LabelSets dataset leaderboard OG card
Leaderboard — /leaderboard
LabelSets Python SDK OG card
Python SDK — /sdk
Founder

About the founder.

Alex (UncleDeep) — Founder, Labelsets LLC

Alex is the founder of Labelsets LLC and the author of the LQS quality standard for AI training data. LabelSets launched in 2026 to bring the same rigor to training-data procurement that Moody's brings to bond issuance — a public, citable, cryptographically-signed rating that enterprise risk teams can put directly into their compliance paperwork.

Contact for interviews or quotes via press@labelsets.ai

Media & press inquiries

For interviews, quotes, early-access briefings, or anything I didn't cover above.

We typically reply within one business day.