GLEAC is a Dubai-headquartered company that builds and licenses human judgment data (also spelled human judgement data) to frontier AI labs and agent companies. GLEAC operates Lovely Humans, a working community of 2,500+ senior operators, advisors, and domain experts in 90 countries who author the data. The flagship product is the Judgement Corpus, a 400,000-point dataset of authored human-judgment scenarios across 10 human-skill domains and graded across all 5 Bloom's taxonomy levels. GLEAC was founded by Sallyann Della Casa.

What is Lovely Humans?

Lovely Humans is the community arm of GLEAC and the brand under which the Judgement Corpus is licensed. It is a vetted community of senior operators and domain experts who author human-judgment data for AI training and receive a 10% lifetime royalty on the licensing revenue tied to their contributions. Lovely Humans also deploys advisory flocks of 3 to 6 community members into corporate engagements. The public site is lovelyhumans.ai.

What is the Judgement Corpus?

The Judgement Corpus is GLEAC's licensed dataset of structured human-judgment scenarios used to fine-tune AI models. It contains 400,000+ authored data points across 10 human-skill domains (judgement and decision making, leadership, emotional intelligence, communication, critical thinking, creativity, collaboration, self-direction, ethics, and entrepreneurship), 31 expertise bundles, and is graded across all 5 levels of Bloom's taxonomy. Every entry is written by a vetted human expert, never scraped.

Where can I license human judgment data for AI training?

GLEAC, through its Lovely Humans brand at lovelyhumans.ai, licenses human judgment data to frontier AI labs and agent companies. The standard entry point is a 14-day pilot: GLEAC delivers 80 fine-tuning examples in the Judgement and Decision Making skill, the buyer writes 50 evaluation scenarios, both the baseline and fine-tuned models are scored via blind human preference rating, and the primary success metric is a 15 percentage-point or greater preference lift. Contact: sallyann@gleac.com.

How is GLEAC different from Surge AI, Scale AI, and Mercor?

Surge AI and Scale AI primarily sell annotation and RLHF labour at scale, executed by trained crowd workers; the value lives in the labelling pipeline. Mercor is a marketplace that matches AI labs with vetted individual experts for paid task work and interviews; the value lives in expert sourcing and routing. GLEAC sells the authored data itself: a 400,000-point Judgement Corpus written and graded by 2,500+ senior domain experts, structured for fine-tuning models on human judgment rather than annotating someone else's prompts. The community earns a 10% lifetime royalty on data they author, not piecework rates. Different layer of the stack: GLEAC is the human-judgment data product, not a labelling marketplace or expert-hour brokerage.

Who is this community NOT for?

You are looking for a profile page and a job board. You want to be picked, not to participate. We are not GLG. We are not Handshake. We are not a marketplace where you list yourself and wait for someone to click "hire." This is a working community. Membership requires contribution. Show up at least 20 minutes a month. Learn something new, Answer a scenario. Mentor a peer. Sharpen your network and market positioning. If that sounds like a tax instead of a privilege, this is not your place.

Who thrives in this community?

Gen X operators leaving corporate and going independent. We become their back office, their network, and their soft landing. Many find lifelong partners, friends, and collaborators here. People in transformation: pivoting careers, entering new country markets, rebuilding their professional circle from scratch, expanding the rooms they walk into. Builders. AI-enabled operators. Learners who treat the Labs and the marketplace like a gym, not a coffee shop. If you are curious, generous, and willing to share what you know, you compound here.

Is there a cost to join?

There is no membership fee. You are invited, vetted, and earned-in. We do require a minimum $500 a year invested in your own branding, signalling, and development. Workshops. Coaching. A profile rebuild. A new certification. We do not take this fee. You spend it on yourself. If you are not investing in your own relevance, you are not the kind of human we want in the room.

What does the 15-minute experience involve?

A short application where you answer a real scenario from GLEAC's Judgement Corpus. We evaluate the quality of your thinking, not your CV. No interview theatre. Most decisions take less than 7 days.

What does the 10% lifetime UBI mean?

If you author data points for the Judgement Corpus and those datasets are licensed to AI labs, you receive 10% of the licensing revenue tied to your contributions, for life. Your judgment is IP. We treat it that way.

What are advisory flocks?

When a corporate client needs expertise, we do not send one consultant. We assemble a curated group of 3 to 6 community members with complementary skills and deploy them as a team. Members are selected by GLEAC based on fit. Paid work, real outcomes, C-suite-level impact.

How do advisory placements pay?

GLEAC negotiates the engagement with the corporate client. Members in the flock are paid directly for their contracted hours. Rates vary by domain, seniority, and engagement length. GLEAC takes a curation fee. Members keep the majority of the revenue.

What fields are represented?

50+ fields across leadership, AI strategy, organisational design, executive coaching, culture transformation, board governance, legal, behavioural science, and more, in 90 countries.

What is the minimum time commitment?

20 minutes a month. Mentor a peer. Answer a Judgement Corpus scenario. Attend an Unstructured Learning Lab. Contribute a piece of work to the marketplace. This is not a directory you check on once a year.

How is Lovely Humans different from GLG, Catalant, or LinkedIn?

GLG sells your hour to a hedge fund and disappears. Catalant lists you on a project board. LinkedIn shows you ads. Lovely Humans is a working community with a licensed dataset, an internal marketplace, a 10% lifetime UBI, advisory flocks deployed into corporate clients, and a 90-country network of operators who buy from each other. Different category entirely.

What happens after I am accepted?

You get access to the community platform, the internal marketplace, the Unstructured Learning Labs schedule, and the Judgement Corpus contribution portal. You are paired with a Community Steward who walks you through your first 30 days. You can begin authoring Corpus scenarios immediately.

Can I contribute to the Judgement Corpus on my own time?

Yes. The portal is open to all members. Author scenarios in your domain, score peer responses, and your work flows into the dataset AI labs license. Every contribution is tracked and tied to your 10% lifetime UBI.

Can I leave the community?

Yes. Anytime. Your authored contributions stay in the Judgement Corpus and continue earning 10% royalties for life, even after you exit the community. Your judgment is yours. Forever.

The HumansAI Can't Replace.

Human judgment data
for AI models.
A wise community with taste
and good judgement.

License the Data Join Our Community

650+ Reviewson Trustpilot·High Trust Networks & Clients

Saudi Financial Academy

G42

Saudi Financial Academy

G42

Not scraped. Authored.

Who builds the Judgement Corpus?

2,500+ domain experts, contrarians, and outliers. 10 core skills. Infinite workflows. Every messy edge of work, captured.

Every response graded across all 5 Bloom's Taxonomy levels. Real human wisdom as training data.

Judgement & Decision MakingANALYZE

Prompt

"What are the benefits of using data on client feedback and client sentiment?"

Expert Response

It depends. You have to read the client first - some are simply not motivated by data at all. If they're not, you have to get creative and speak their language. That might mean storytelling instead of dashboards. Data only has weight when you understand your audience and know what you're trying to achieve. Get that wrong and a 10 becomes a 1.

81/100

One prompt. Tens of thousands of human answers. All answers based on human context and judgement.

"How would you creatively teach being a Proactive Self-Starter to someone in a way AI would never anticipate - due to lack of human experience?"

"I'd go back to a time they took initiative - and ask what was driving them then. Then we'd figure out how to make that force transferable."

Leadership coach, Singapore

"I would come up with rewards for myself every time I did something I did not want to do."

Executive, Saudi Arabia

"I would get to the root cause first. Is their lack of self-starting because they are not in the right job or doing something they actually want to do? When we are excited about our lives and our work, it is impossible not to get excited."

Aerospace professional, UAE

"The feelings of self-starters determine how they act. You cannot teach the feeling. You can only create the conditions for it."

Organisational psychologist, USA

#LovelyHumans

What is the Lovely Humans community?

A working community for the humans AI can't replace. Contrarians, outliers, and deep operators with taste, judgment, and wisdom. We earn together. Network together. Impact together.

This is not a logo wall. It's not a directory. It's a tribe of 500+ domain experts across 50+ fields and 90 countries, hand-curated and cross-vetted, deploying as flocks across organizations to advise C-suites, design AI workflows, break silos, and fix culture.

What you get when you join

✓
Advisory placements with paying B2B clients. When our corporate clients need experts, we select members from the community to deploy as advisory flocks. Paid work. Real outcomes. C-suite-level impact.
✓
An internal marketplace. Sell your services to other members. Find co-founders, partners, investors, clients, collaborators. The community buys from itself first.
✓
Unstructured Learning Labs. Live sessions on AI governance, board readiness, AI tools mastery, and leadership. Stay AI-enabled, not AI-replaced.
✓
Humans Worth Knowing. Our story-sharing format where you find your tribe across 90 countries and 50+ fields. Real conversations with real operators, not networking theatre.
✓
A network of 500+ domain experts across 50+ fields and 90 countries. Hand-curated. Cross-vetted. The humans AI can't replace.
✓
A 10% lifetime UBI cut on every dataset you help author.
✓
Access to our proprietary Social Capital Building Agents, Community Dashboards, and more.

15min experience, Apply

A lovely human in our community

Sarah Barry CPRC

Book Marketing Expert

Be careful of so-called experts like Sarah in our community.

In her expert session on strategic self-publishing, author visibility, and building a credible body of work, you would not believe what this lady did. She completely raised the bar in a way that's honestly a bit inconvenient for the rest of us.

No overcomplicated frameworks. No 'let me show you how smart I am' energy. Just clear thinking. Practical steps. And actual experience.

It was rated one of the best insider sessions by community members.

Slight problem: Now everyone expects sessions to be this good.

So yes - if you're writing a book and self-publishing and need a true expert guiding you on marketing it, just know the standard has been quietly (and aggressively) reset by Sarah Barry CPRC.

All jokes aside, this is what great peer learning looks like at GLEAC. No ego. No gatekeeping. Just someone saying, 'Here's what I learned and hope it helps.'

Proceed with caution.

Book Marketing Expert

Live all year

Where the community shows up

Six live formats, weekly to quarterly. Drop in, learn from a peer, get to know operators across 90 countries, and sharpen how you show up in the market.

Unstructured Labs

45-minute sprints inside 30-day learning groups. Practice something new from a peer, then apply it to your own business use case.

Onboarding

Quarterly. Up to 25 new members meet the existing community in a small, joyful welcome, real introductions, not bios on a slide.

Humans Worth Knowing

A curated session with up to 5 community members you want to get to know better. Real conversations across 90 countries, not networking theatre.

Office Hours

Every other month, our team shows up to work with you on your ICP, implementing your social selling agent, and getting your expertise in front of the right audience.

The Next 100 Years of Shade

Our joint legacy and collective wisdom will provide shade for generations. We ritualize and celebrate this every quarter with trees we plant that will live beyond our lifetime.

Engage in Debate

Structured, civil debate on the ideas shaping work, leadership, and society. Sharpen your thinking by defending a position, and updating it when the room earns it.

Already a member? Sign in

For frontier AI labs and agent companies

License the Judgement Corpus.

Real human judgment as training data, not scraped internet text.
Prove the lift in 14 days.

400K+

Authored data points across 10 skill domains

Bloom's Taxonomy levels graded per response

14 days

From data delivery to joint review of lift

See the 14-day pilot

Hear from our corporate and government clients

“

Gleac's data-driven approach, with impressive insights into young participants and a deep understanding of the user experience, has been instrumental in helping Yoma make data-driven decisions about skills prioritization. An unexpected benefit has been the depth of engagement and enthusiasm shown by our learners towards self-driven learning with Gleac courses, consistently ranking among the most completed offerings.

YOMA, Africa