Metaview just got a whole lot smarter, now with Multi-source
By the time a candidate is in their fourth round, the recruiter is the only person who has been in every room. The phone screen, the technical, the panel debrief: each one produced its own summary, and the work of stitching them into a hiring decision lands on the human.
Single-call AI was a good start for the category. It delivered a clean transcript, a structured summary, time back per interview. But the loop is the unit of work, and the loop has always been somebody's mental tab-juggling exercise.
Multi-source moves the stitching upstream. Our AI reads across the loop: every interview transcript plus the JD, the resume, the rubric, and the internal docs that define what "great" looks like. Together, they surface the brief the next interviewer needs.
AI is only as good as the context you give it
Most AI notetakers analyze one conversation at a time. That was the right starting point for the category. Get a clean transcript and a structured summary out of a single call, ship that as AI in recruiting 1.0, move on.
But across a hiring loop, single-call AI leaves a lot on the floor. Missed signals. Shallow summaries. Repeated questions from one interviewer to the next, because nobody told them what the first round already covered.
Multi-source treats the loop as one body. By analyzing every conversation, every doc together, our AI mirrors your bar, adapts to your process, and continues to get sharper with every call that lands.
How Multi-source works
Before the mechanics, here is the 60-second walkthrough of Multi-source in the product.
Five inputs feed the engine. Conversations come from the calls Metaview already captures through Notetaker on Zoom, Teams, Meet, phone, and in-person. JDs and resumes flow through your ATS. Interview rubrics and internal docs upload via the workspace doc layer.
The analysis layer then reads the corpus through the lens of your hiring bar. What your rubric defines as "strong," what your prior interviews uncovered, what's still missing. The output isn't a transcript summary anymore. It's a hiring brief.
None of the source content moves outside your existing data boundary. Every conversation and document stays inside the workspace, governed by the same SOC 2 and ATS data agreements you already have.
How customers use it
We designed Multi-source around the workflows that already eat the most recruiter time. The four use cases below are the ones early customers reached for first, and the ones that paid back fastest.
Send hiring managers into the panel with structured notes on what's been covered, what's still unclear, and where to dig deeper.
Compile a full candidate pack from resume, interview insights, and role alignment so the panel walks in informed, even when the CV is thin.
Synthesize every interview on every shortlisted candidate in one place to see how they stack up against the role and each other.
Spot the questions a round missed and the rubric dimensions the panel didn't probe, before the loop closes.
Because Multi-source treats the corpus as one body, the same engine produces all four outputs from the same underlying analysis. The team picks the view that matches the moment.
What that means in practice
Most of the workflow Multi-source replaces was invisible. A recruiter watches prior calls back at 1.5x. A hiring manager scrolls the JD in another tab. An interviewer opens the rubric for the third time to remember what "communication: 4" means.
The cost showed up as longer days and shallower debriefs, not as a missed step in the process.
- Re-watch prior interview recordings to know what's been asked
- Stitch JD, resume, and rubric across tabs for every debrief
- Hiring manager opens the panel with a fresh-eyes summary, not a brief
- Open the next interview with the prior context pre-analyzed
- The hiring manager pack writes itself from the corpus
- Side-by-Side comparisons exist the moment a second candidate runs the loop
The shift is concrete. The brief writes itself, and the recruiter walks into the panel with twenty to thirty minutes of cross-tab work already done.
Built with the best
We love co-building with the recruiters, RecOps leaders, and TA pros in our 10x Recruiting community, the people pushing the edge of what AI can do in hiring.
For Multi-source, three customers shaped the earliest use cases: Nate Wylie at Infinite Talent, Samy Aumar at Qonto, and Margot Gorrin at Mercury.
Customers using Multi-source for higher-volume hiring describe the same mindset shift.
We've completed over 1,900 calls using this platform, saving 77 full workdays. We're not just automating note-taking, we use the multi-source feature so each interviewer goes in unbiased but informed enough to cover new ground."
Multi-source is live in every workspace today. You can start free on the next req, or book a demo to see the brief generate in real time on your own loop.
Frequently asked
What sources can Multi-source ingest?
Setup is self-serve via Admin Panel for most ATSes and doc sources. Conversations come through Metaview Notetaker. JDs and resumes flow via your ATS integration. Rubrics and internal docs upload through the workspace doc layer. If your stack isn't listed yet, ask your Metaview contact about the roadmap.
Do I need to opt in, or is Multi-source on by default?
On by default in every workspace from launch. Existing customers were auto-enabled. New accounts get Multi-source ready to use the moment Notetaker captures a first conversation. If a workspace ever needs to scope it down, the toggle lives in Settings > AI Notes preferences.
How does Multi-source affect candidate privacy and data residency?
Every conversation and document ingested stays inside the workspace's existing data boundary; nothing crosses tenants. SOC 2 Type II, GDPR, and your ATS data agreements continue to apply. Our security page is the source of truth for the full compliance scope.
What if my team only has one or two prior conversations on a candidate?
Multi-source works on whatever's available. A single call plus the JD, resume, and rubric is already four sources, and the analysis runs the same way. The output sharpens as more interviews land, but the first-pass brief is value-positive from call one.
How does Multi-source compare to Multi-Source Summaries on the Reports surface?
Multi-source is the upstream ingestion and analysis engine. Multi-Source Summaries on Reports is one of the user-facing views the engine powers, the offer-prep brief. One is the engine. The other is a view.
Bring Metaview into your hiring stack.
Live notes, structured scorecards, and ATS sync - set up in under 10 minutes.
