Purpose
Gradientflo is the second application on Smaya, and it covers everything after the hire: onboarding, assimilation, ramp, role fit, development and performance. At its core, Gradientflo teaches. Onboarding and project training are delivered by AI tutors — conversational agents that carry your project, team and company knowledge to a new hire and coach them through the first ninety days, over voice and avatar rather than a portal to read. Everything else in the application exists to support that: checking in on the hire, spotting when they are struggling, and re-testing whether the role is working. Exterview decides who joins. Gradientflo gets them productive, and tells you honestly when the decision was wrong.Status: Upcoming. Gradientflo is a separate application on Smaya, and
nothing on these pages is available today. They are published so customers and
partners can see the direction and influence the sequence.
The problem it solves
A hire is evaluated with rigour, against explicit criteria, with evidence and a defensible record. Then they start — and all of it is thrown away. The manager receives no structured account of what the person was strong and weak at. The 30/60/90 depends on whether that manager has the bandwidth to remember it. When someone struggles, nobody can say whether the interview missed it or the job changed. And when they leave in month seven, the exit interview is the first structured conversation anyone has had with them since the offer. Gradientflo closes that gap by carrying the hiring evidence forward and continuing to evaluate against it.What it covers
Onboarding & training
An AI tutor that teaches a new hire their project, team and organization from day one and answers from your own material — so ramp does not depend on a colleague being free.
Assimilation
A reliable check-in rhythm at 30, 60 and 90 days that runs regardless of
manager bandwidth, and leaves a record that it happened.
Role fit
Re-evaluation at day 60 and 90 against the criteria the person was actually
hired on — closing the loop rather than starting a new one.
Development
A gap turned into a specific plan, mapped to learning the organization already
licenses.
Performance
Review and calibration against the same skill language used at hire.
Exit
The exit as a lifecycle event, with the interview itself run on Exterview’s interviewing capability.
What it deliberately is not
Gradientflo is not an HR system of record, and that boundary is absolute.
The test: if the output is a record, it belongs in your HR system. If the output is a judgment backed by evidence, it is ours. That is why exit interviews are in scope and offboarding checklists are not.
Why it is one platform with Exterview
Because the evidence is the same evidence. A candidate evaluated against a skill, an employee developed against that skill, and a review conducted against that skill are one data model, not three. Built separately, you forfeit the only durable advantage either product has. See The Talent Graph.Who buys it
The CHRO or Chief Learning Officer originates it. The CIO holds the veto — because Gradientflo is not evaluated as an HR application. It is evaluated as a set of autonomous agents operating on employment data inside your tenant. That is why its governance is not a later hardening pass. Notice, consent, bounded authority, human approval and a complete record are in the first release, not the third.FAQs
Do we need Exterview to use Gradientflo?
Do we need Exterview to use Gradientflo?
They are separate applications on one platform and can be adopted in either order. The combination is worth more than either alone, because the hiring evidence is what makes post-hire evaluation continuous rather than a fresh start.
Is this a learning management system?
Is this a learning management system?
No. Gradientflo connects to the learning your organization already licenses
and maps development gaps to it. We produce the judgment about what someone
needs; we do not host the courseware.
Will AI decide performance outcomes?
Will AI decide performance outcomes?
No. Role fit and development output is evidence prepared for a manager.
Judgments about a person’s performance are made by a named human, as they are
everywhere else on the platform.
When will it be available?
When will it be available?
It is Upcoming. See Roadmap for the sequence. Releases are sequenced by readiness, and we do not publish dates.
Related
- Gradientflo Agents. The ten agents.
- The Talent Graph. What connects it to Exterview.
- Roadmap.