Plugging Async Video Interviews Into Your ATS Without Breaking Your Pipeline

TL;DR

Async video interviews belong inside your ATS workflow, not next to it. The right integration pattern is stage-triggered: when a candidate moves into a screening stage in Greenhouse, Lever, Ashby, or Workday, the video interview invite fires automatically, and the completed interview writes back a score, a summary, and a deep link to the recording. Done well, recruiters never leave the ATS to make a decision, candidates never receive duplicate emails, and your reporting stays in one system. This guide covers the integration patterns for each major ATS, the data that should flow in each direction, and the mistakes that quietly corrupt pipelines.

Why Do Video Interview Integrations Break Pipelines in the First Place?

Most hiring teams do not fail at async video interviewing because of the video part. They fail at the plumbing. The typical broken setup looks like this: recruiters export candidates to a CSV, upload them to a separate video tool, review recordings in that tool, then manually retype outcomes into the ATS. Every manual hop introduces three problems:

• Stale pipeline states. A candidate is rejected in the video tool but still shows as active in the ATS, so a second recruiter schedules them for the next round anyway.
• Split reporting. Time-to-hire, pass-through rates, and source quality live half in the ATS and half in the interview tool. Neither number is trustworthy.
• Candidate confusion. Two systems send emails on their own schedules. Candidates get an interview invite after they have already been rejected, which is a brand problem, not just a data problem.

The fix is not more discipline. It is an integration where the ATS remains the single source of truth and the video interview platform behaves like a stage inside it. With the average time-to-hire now at 44 days (Gem, 2025) and 60% of candidates abandoning slow processes (iCIMS, 2025), you cannot afford workflow friction that adds days of manual handling to every requisition.

What Should a Good ATS Integration Actually Do?

Strip away the vendor language and a proper integration comes down to four flows. If a video interviewing platform cannot do all four, you will be doing manual work forever.

• Trigger on stage change. When a recruiter moves a candidate into the "Video Screen" stage in the ATS, the invite goes out automatically. No exports, no copy-paste, no separate send button.
• Sync candidate data one way, cleanly. Name, email, requisition, and locale flow from the ATS to the interview platform. The ATS record is never edited by the video tool, so there is exactly one master record.
• Write results back where decisions happen. When the async interview is completed and scored, the AI score, a structured summary, and a deep link to the recording appear on the candidate profile inside the ATS. Reviewers should be able to act without switching tabs.
• Mirror status in real time. Invited, started, completed, expired, advanced, rejected. Both systems agree on where every candidate stands at every moment, typically via webhooks rather than nightly batch jobs.

This matters more as volume grows. Structured, consistent screening is twice as predictive of job performance as unstructured evaluation (Schmidt & Hunter, 1998), but structure only survives at scale when the workflow is automated end to end.

How Does the Integration Pattern Differ by ATS?

Greenhouse

Greenhouse is the most integration-mature of the four. The standard pattern uses the Harvest API plus a custom interview stage: you add "Async Video Interview" as a stage on the interview plan, and moving a candidate into it triggers the invite. Results return as scorecard-style notes and links on the candidate profile. Two practical tips:

• Map the video interview to a dedicated stage rather than piggybacking on "Application Review." Pass-through reporting per stage is one of Greenhouse's strengths, and you want the video screen measured on its own.
• Use Greenhouse webhooks for rejection events so an in-flight video invite is cancelled the moment a candidate is rejected elsewhere in the process.

Lever

Lever thinks in pipelines and opportunities rather than interview plans. The clean pattern is stage-based triggers through the Lever API: entering a designated pipeline stage fires the invite, and completed interviews come back as notes with links plus tags for score bands (for example, a tag for candidates scoring in the top quartile). Tags matter in Lever because recruiters filter and build views on them, so a score band tag turns AI scoring into something recruiters can actually slice their pipeline by.

Ashby

Ashby's API is modern and its interview plan model is flexible, which makes it the easiest of the four for bi-directional sync. The pattern mirrors Greenhouse: a dedicated interview stage triggers the invite, and results write back to the candidate timeline. Because Ashby's analytics are unusually strong, make sure the completion timestamp and score come back as structured fields, not just free text, so they show up in Ashby's native reporting.

Workday

Workday is the enterprise case and deserves honest expectations. Integrations run through Workday's integration framework and typically involve IT, security review, and a sandbox tenant. Plan for weeks, not days, and involve your Workday administrator early. The upside: once the integration is live, Workday's business process framework can route video interview outcomes into approval chains and compliance workflows that smaller ATSs cannot match. For enterprise teams, insist on SSO alignment and a sandbox test cycle before anything touches the production tenant.

What Data Should Flow Back Into the ATS?

Sending "interview complete" is not enough. The write-back payload determines whether reviewers can make decisions inside the ATS or still have to open a second tool. A complete write-back includes:

• The AI score and per-question breakdown. StormInterview scores every response against the role's criteria, so the ATS profile shows not just "7.8 overall" but how the candidate did on each competency.
• A structured summary. Three or four sentences a hiring manager can read in ten seconds, attached as a note.
• A deep link to the recording. One click from the ATS profile opens the exact interview, already scoped to the right reviewer permissions.
• The review decision. When a recruiter swipe-reviews candidates in StormInterview (advance or reject, one decision per swipe), that decision moves the candidate in the ATS automatically. Screening 200 candidates in 30 minutes only helps if the results land in the system your team actually works in.
• Consent and retention metadata. Where the recording is stored, when it expires, and what the candidate consented to. Your compliance team will ask, and the answer should live in the ATS audit trail.

Which Integration Mistakes Quietly Corrupt Pipelines?

These are the failure modes we see most often in teams migrating from a manual setup:

• Duplicate emails. Both the ATS and the video platform send an invite. Decide which system owns candidate communication per stage and disable the other. One owner per message, always.
• Orphaned invites. A candidate is rejected or withdraws in the ATS, but the video invite stays live. Wire rejection webhooks to auto-expire invites.
• Free-text score dumping. Scores pasted into a comment field cannot be filtered, reported on, or audited. Insist on structured fields or tags.
• Testing in production. Every major ATS offers a sandbox (Greenhouse, Ashby, and Workday explicitly; Lever via a test environment). Run at least one full candidate journey, invite through decision through write-back, before going live.
• Ignoring rate limits. Bulk-inviting 500 candidates can hit ATS API limits and silently drop invites. A production-grade integration queues and retries; ask your vendor how theirs handles a burst.
• Skipping the locale field. If you hire in multiple languages, the candidate's locale must flow from the ATS so interview instructions and questions arrive in the right language automatically.

How Do You Roll Out an ATS Integration Without Disrupting Live Requisitions?

You do not need a big-bang cutover. The proven sequence:

• Week 1: Sandbox validation. Connect the sandbox, run three test candidates end to end, and verify every status transition appears correctly on both sides.
• Week 2: One pilot requisition. Pick a live, high-volume role. Run the integrated flow with one recruiter who reports friction daily.
• Week 3: Fix the mapping. The pilot always surfaces mapping decisions: which stage triggers the invite, who owns rejection emails, which score threshold auto-advances. Lock these down in writing.
• Week 4: Expand by team. Roll out team by team rather than all at once, and keep the manual path available for one more cycle as a fallback.

Measure before and after: invite-to-completion time, screening stage duration, and recruiter hours per hire. Teams that connect async video screening to their ATS typically see the screening stage compress from one to two weeks of phone-screen scheduling to two or three days, since video review is roughly 6x faster than phone screening (interviewstream, 2025) and 42% of scheduling-related dropout (Cronofy, 2024) simply disappears.

How StormInterview Handles ATS Integration

StormInterview is built to sit inside your ATS workflow rather than beside it. Stage changes in your ATS trigger interview invites automatically, candidates record async video responses on their own schedule, AI scoring evaluates every answer against your criteria, and recruiters swipe-review the results in minutes. Every decision writes back to the ATS as structured data: score, summary, recording link, and outcome. Your pipeline stays in one system, your reporting stays accurate, and your team screens 200 candidates in 30 minutes without ever exporting a CSV.

Start a free trial of StormInterview and connect your first requisition this week, or talk to our team about your Greenhouse, Lever, Ashby, or Workday setup.

Frequently Asked Questions

Does an ATS integration replace my existing interview stages?

No. It adds a video screening stage to your existing interview plan. Later stages (live interviews, offers) stay exactly as they are. The integration only automates the invite, the score write-back, and the stage movement for the video screen itself.

Which ATS is easiest to integrate with async video interviews?

Greenhouse and Ashby are typically fastest because of their mature APIs and dedicated interview stage models; most teams are live within days. Lever follows closely using stage-based triggers and tags. Workday takes the longest because integrations pass through IT and security review, so plan for a multi-week timeline with sandbox testing.

Do AI interview scores show up inside the ATS?

With a proper integration, yes. Scores should arrive as structured fields or tags on the candidate profile, together with a summary and a deep link to the recording, so reviewers can decide without leaving the ATS and analytics teams can report on score bands natively.

What happens if a candidate is rejected in the ATS while a video invite is still open?

A well-built integration listens for rejection webhooks and expires the invite immediately, preventing the candidate from receiving further emails or completing an interview that no one will review. Ask any vendor to demonstrate this specific flow before you sign.

Is a two-way ATS integration safe for candidate data?

Yes, when scoped correctly. The ATS remains the master record, the video platform receives only the fields it needs (name, email, requisition, locale), and consent plus retention metadata is written back to the ATS audit trail. Review vendor SOC 2 status and GDPR data processing terms as part of procurement.