LoginSign Up
← Back to all sparks
T

Teamtailor

HR
Velocity6.3

Teamtailor's Co-pilot is quietly automating the full recruiting pipeline, one trigger at a time.

atsrecruiting-automationco-pilotagentic-workflowsmcp-integrationcandidate-pipeline
◆Current state
Teamtailor has expanded well beyond a standard ATS into a platform where AI handles candidate intake, data collection, screening, and stage routing with minimal recruiter input. The Co-pilot layer now covers agentic workflows — requesting missing info, re-screening at any stage, detecting prompt injection in resumes — alongside a fresh MCP integration that opens the platform to external AI orchestration. The candidate surface has also broadened: WhatsApp messaging, group meeting improvements, and onboarding video support show a product filling in every communication and handoff gap.
◆Where it's heading
The pattern across the last six months is consistent: Teamtailor is building toward a recruiting pipeline where the recruiter sets rules once and Co-pilot executes them end-to-end. Each release either expands what Co-pilot can trigger autonomously (information requests, screening re-runs, smart moves) or opens the platform to external agents via MCP. Onboarding and communication features are expanding in parallel, suggesting the product ambition is the full employee lifecycle from application to first week — not just the hiring funnel.
◆Prediction
The most likely next move is Co-pilot-driven interview scheduling — the last high-friction manual step before offer — possibly using the MCP channel to coordinate with calendar tools like Google Calendar or Calendly without leaving the platform.

◆Recent moves

  1. 1d ago

    Group meetings just got better ✨

    Group meetings now support explicit decline responses from candidates, giving recruiters accurate attendance status (invited / accepted / declined / no response) instead of treating silence and rejection as identical. Message templates, file attachments, and automatic calendar entries for interviewers round out what was previously a stripped-down scheduling flow. This is steady UX work — filling in gaps that made group meeting coordination unreliable — rather than a directional move.

    View source ↗
  2. 7d ago

    Add videos to Onboarding dashboards, and a clearer onboarding checklist

    Video support lands in onboarding dashboards, letting companies embed welcome videos or team introductions directly on the new-hire experience. The to-do sorting and grouping update is minor UX polish. Together these extend Teamtailor's onboarding surface without adding new automation; they're content-enrichment features, not workflow changes.

    View source ↗
  3. 20d ago

    Missing candidate info? Co-pilot’s on it.

    ⚡ SPARK

    Co-pilot can now proactively identify gaps in a candidate's profile — missing resume, unanswered form questions, unfilled custom fields — compose a targeted questionnaire, send it to the candidate, and route the responses back into screening and smart-move triggers without recruiter intervention. This closes the last manual loop in automated candidate processing: the platform can now intake, screen, collect missing data, and advance candidates end-to-end.

    View source ↗
  4. 26d ago

    Co-pilot now detects hidden instructions in resumes

    Co-pilot now runs a prompt injection scan on every uploaded resume, flagging hidden text (color-matched, transparent, or microscopic) that could manipulate AI assessments. Detected instructions are ignored and logged, with a badge indicating clean or flagged status. This is a meaningful defensive addition as AI-scored hiring becomes more common, but it's a guard rail on an existing feature rather than a new capability direction.

    View source ↗
  5. 2mo ago

    Connect your AI assistant to Teamtailor with MCP

    ⚡ SPARK

    Teamtailor ships an MCP add-on that lets external AI assistants — Claude, ChatGPT, or any MCP-compatible client — read and write directly to jobs, candidates, stages, and custom fields. This is the most structurally significant release in recent history: it opens Teamtailor as a data and action layer inside whatever AI orchestration workflow a company already runs, rather than requiring all AI work to happen inside the platform's own Co-pilot.

    View source ↗
  6. 2mo ago

    Stage-by-stage screening and a redesigned trigger view

    Screening criteria can now run as triggers at any pipeline stage, not just the inbox, with different criteria per stage. Template inheritance and job-copy carry-over remove per-job setup friction. This meaningfully expands where the AI screening layer operates, enabling multi-stage qualification rather than a single intake gate — consistent with Co-pilot's broader expansion into pipeline automation.

    View source ↗