SkillsDBvsGloat

A health insurer launches a ten-week project to automate prior-authorization reviews. It needs five people at Level 3 in Python and Level 3 in clinical coding. The marketplace posts the gig and returns forty strong matches, drawn from resumes, past projects, and profile skills. Every one of them is a plausible fit.

In SkillsDB, the same forty people look different. Seven are assessed at Level 3 or above on both skills, rated by their managers against anchored levels. Twelve are at Level 2 in Python and would spend the first month learning. The rest have Python on their profile from a course they took, not from work anyone has assessed.

The project lead staffs from the seven, and the twelve get development plans aimed at the next project, not this one. Nobody gets pulled off their day job for a gig they cannot yet deliver, and the marketplace's next match runs on verified levels.

Gloat built the internal talent marketplace category and still defines much of it. Its platform matches employees to projects, gigs, mentorships, and open roles from a knowledge graph of people, skills, roles, and relationships, and it now positions itself as an agentic HR platform: career agents that infer skills from resumes and completed work, workflow agents that automate HR processes, and strategic agents for planning. For a large enterprise that wants to unlock lateral movement and redeploy people quickly, that machinery is real and mature.

The inference engine is the point. Gloat's public materials describe agents that add skills to a profile after a project ships or a resume is uploaded, then match on semantic embeddings rather than keyword overlap. Combined with integrations into Workday, SAP SuccessFactors, Oracle HCM, and Microsoft Teams, and a security posture that publicly documents SOC 2 Type II and ISO 27001, 27017, and 27018, Gloat is a credible system for moving talent around a company of tens of thousands.

Inferred vs assessed. Inference is the fastest way to get from an empty profile to a plausible one, and SkillsDB will ingest exactly that kind of data as a starting point. Inference gets you started. Assessment tells you where you actually stand. An inferred skill cannot tell you whether the person can perform it today, at what level, or who confirmed it. SkillsDB verifies through structured skills assessment: a required proficiency level per role, an assessed level per person, and who rated it, by what method, and when.

Fit scores vs gaps. A fit score compares a profile to an opportunity. A gap compares an assessed level to a required one on a defined scale. SkillsDB turns the gap into work: gap analysis per person, role, and team, development plans tied to the specific skills that are short, and scorecards that compare self, manager, and benchmark ratings so inflated self-ratings surface instead of driving matches.

Mobility layer vs the record underneath. Gloat sits on top of whatever skills data exists. SkillsDB is the record underneath: competency frameworks in your own terms, assessment, the skills matrix, certifications, development plans, and career pathways from the same person-to-skill records. Most organizations running a marketplace still need that record, and many organizations running SkillsDB have no need for a gig marketplace.

Security posture. Both vendors take this seriously. SkillsDB holds a SOC 2 Type II attestation, independently audited by Johanson Group LLP, and is GDPR compliant, with SSO and SCIM provisioning. Gloat publicly documents SOC 2 Type II and several ISO certifications. Ask each vendor for the reports themselves.

Choose Gloat if…

• Your priority is internal mobility at enterprise scale: gigs, projects, and lateral moves across tens of thousands of people.
• You are comfortable matching on inferred profiles because the opportunities are exploratory, not high-stakes.

Choose SkillsDB if…

• You have staffed a project from marketplace matches and watched the first month go to learning the basics. You need assessed levels before the match, not after.
• Managers are asked to release people for gigs and want to know the person can deliver before they say yes.
• People who are one level short should get a plan to close the gap, not just a lower fit score.
• Your security review expects a SOC 2 Type II report and your IT team expects SCIM provisioning, and you want both answered in week one.

The complementary pattern is straightforward. SkillsDB holds the verified record: the skills library, the proficiency scale, assessments with dates and assessors, and the role requirements from your job architecture. A marketplace consumes that record to make its matches more accurate, because a match against an assessed Level 4 means something a match against an inferred tag does not.

Practically, that means agreeing which system is the source of truth for a skill and its level, syncing assessed levels outward rather than inferred tags inward, and treating marketplace activity as a signal to trigger a reassessment rather than as proof of competence. For a longer treatment of the category boundary, see Talent Marketplace vs. Skills Platform.

• Show me the top ten matches for this project. Which of them are at the required level on every skill, and how do you know?
• How does a skill get onto an employee's profile, and can the employee or their manager see why it is there?
• Is proficiency expressed on a defined scale with behavioral anchors, and does a level mean the same thing across skills?
• Who assessed a skill, by what method, and is the assessor and date recorded?
• When someone is one level short of an opportunity, what happens next? Is a development plan created from the gap?
• Can I see the SOC 2 Type II report itself, and where is the data hosted?