Elicivo replaces rigid scoring boxes with natural AI conversations. Your reviewers talk; the intelligence happens behind the scenes — fully logged, fully auditable.
Admins build once. AI conducts, scores, and consolidates. Every step is observable and adjustable.
Use the visual workflow builder to connect interview stages, rubric templates, alignment settings, and branching conditions. Save and reuse across cycles.
Reviewers receive a link and have a natural conversation. The agent follows your rubric invisibly — no forms, no checkboxes. Sessions are resumable if interrupted.
A separate aggregation agent consolidates all sessions, flags discrepancies, computes an overall score, and generates a narrative report — with a full audit trail.
Built for admins who need control and reviewers who need simplicity.
Reviewers have a natural chat. The AI steers toward your criteria without them ever seeing a rubric or a score input.
Drag-and-drop pipeline designer. Connect interview stages, rubric templates, conditions, and aggregation nodes as a directed graph.
Set per-employee review intervals. The dashboard surfaces overdue employees and triggers new cycles against the right workflow automatically.
Participants can see their per-criterion scores and request adjustments. Large adjustments trigger an agent-validation step before acceptance.
Each interview node can be scoped to SELF, PEER, or MANAGER assignments — with conversation framing and context automatically adapted for each.
Configure per-interview topic blocklists. The agent redirects off-topic or sensitive conversations without passing those messages to the scoring step.
Configure each interview and aggregation stage with an explicit alignment — no hidden defaults, no opaque AI behaviour.
Surfaces nuance and context. Gives the benefit of the doubt and weights positive evidence more heavily in aggregation.
Probes strengths and development areas equally. All reviewers weighted the same in aggregation. The default for most cycles.
Challenges vague answers and demands concrete evidence. Aggregation holds a high bar and discounts unsupported high scores.
A SHA-256 hash-chained audit trail covers every score change, session event, aggregation step, and admin edit — making retrospective tampering detectable.
Each event stores the hash of the previous event. A verification endpoint replays the chain and confirms no records were altered.
Every event carries an actor type — AGENT, PARTICIPANT, ADMIN, or SYSTEM — so you always know who or what caused a change.
Browse events across your entire organisation, or drill into a single employee review's audit tab to trace the exact timeline.
Join organisations moving to AI-conducted interviews. No forms. No boxes. Just conversation — with every score justified and every action logged.