Evaluation report · Jul 21, 2026
Mark Ellis
Export PDF
Your decision
Recorded beside the draft score; it never replaces it.
Still your recruiter’s call.
Summary
Drafted by HireanexMark backed his headline claim with numbers: a cache on the price lookup took checkout p95 from 840 ms to 310 ms, measured on the same dashboard over two weeks. He described an incident he owned end to end, rolling back before looking for the cause. The thinner area is work with product and design, where he usually joined once the designs were final.
Against the role’s requirements
Must-haves first · in the candidate’s words · the time opens the recording
-
Cuts API latency with measured resultsMust-have · 9/10
DemonstratedShown
We moved the price lookup into a cache — p95 went from 840 ms to 310 ms.
at 10:12at 6:14 -
Keeps services covered by tests and CIMust-have · 8/10
DemonstratedShown
Every pricing change ships with a contract test in CI.
at 7:26at 9:05 -
Owns production incidents end to endMust-have · 8/10
DemonstratedStated
…we rolled back first and only then went looking for the root cause.
at 12:31 -
Works closely with product and designNice to have · 6/10
PartialStated
I usually joined once the designs were final.
at 17:58at 15:48 -
Mentors other engineersNice to have · 8/10
DemonstratedShown
I paired with two new engineers through their first on-call week.
at 23:40at 18:36
Highlights
Strengths
- Measures before and after: every performance claim came with a number.
- Clear incident habits: stabilise first, then find the cause.
Worth asking next
- A time he shaped a design before it was final.
- How he decides a rollback is safer than a forward fix.
In their own words
Verbatim from the transcript
I’d rather ship a smaller change I can measure than a big one I can’t.
at 11:20at 8:02On call is where you learn what the system really does.
at 24:10at 19:20
Interview recording
Transcript
- Maya ChenRecruiter10:05
You said you cut p95 latency on the checkout API. Walk me through one specific change you made and how you measured it.
- Mark EllisCandidate10:12
We moved the price lookup into a cache — p95 went from 840 ms to 310 ms. We measured it on the same dashboard over two weeks.
- Maya ChenRecruiter10:56
What was the actual bottleneck?
- Mark EllisCandidate11:03
We were re-pricing the whole cart on every request. Caching the price lookups took that off the hot path.
- AI InterviewerAI11:52
Tell me about a production incident you were on the hook for. What did you check first, and how did you decide when it was safe to call it resolved?
- Mark EllisCandidate12:08
The first thing I did was pull up the error-rate dashboard and check whether the spike lined up with the deploy.
- Mark EllisCandidate12:31
It did, so we rolled back first and only then went looking for the root cause.
- AI InterviewerAI12:34
You mentioned rolling back before the error rate settled. What told you a rollback was safer than a forward fix — and how long did you wait before calling it resolved?
Draft score
Human-ledStandardof 100High confidence
Every must-have was asked and answered with an example.
4 demonstrated, 1 partial
Requirements evidenced: 5 of 5 · Drafted against your job description and your questions · depth set by the interviewer
Dimension scores
- Technical Knowledge9/10
- Problem Solving8/10
- Communication8/10
- Experience Relevance8/10
- Culture Fit7/10
Integrity check — 1 signalIntegrity check — no signals flagged
Clear- Tab switches 1×Measured
- Tab switches noneMeasured
- Reply pattern naturalInferred
- Camera real (not virtual)Measured
These signals do not decide the outcome.
Your notes
Saved with Mark’s profile · team-only
Strong on measurement. Ask about design collaboration in the next round.
Draft scores are prepared from the transcript and your rubric using AI, then reviewed by your team. A scorecard is never a hiring decision.
Request a demoThe page a hiring manager or a client opens from your link, with no login. Your decision, the integrity check, your notes and the dimension scores are not on it, and the draft score appears only if an admin turns it on.
Mark Ellis
Download PDFSenior Backend Engineer · Req 4471
Interviewed July 21, 2026 · By Maya Chen, live on video · 28 min, recordedBy Larkfield Talent Partners’ AI interviewer · 21 min, recorded
Mark is the strongest of the three I spoke to this week. He backed the latency claim with numbers and owned an incident end to end. Design collaboration is the one area I’d probe.
Maya Chen · Larkfield Talent PartnersMust-haves: 3 demonstrated
Against the role’s requirements
Must-haves first · in the candidate’s words
Must-have
- Cuts API latency with measured resultsDemonstrated
We moved the price lookup into a cache — p95 went from 840 ms to 310 ms.
at 10:12at 6:14 - Keeps services covered by tests and CIDemonstrated
Every pricing change ships with a contract test in CI.
at 7:26at 9:05 - Owns production incidents end to endDemonstrated
…we rolled back first and only then went looking for the root cause.
at 12:31
Good to have
- Works closely with product and designPartial
I usually joined once the designs were final.
at 17:58at 15:48 - Mentors other engineersDemonstrated
I paired with two new engineers through their first on-call week.
at 23:40at 18:36
Interview recording
28:1421:06In their own words
Verbatim from the transcript
I’d rather ship a smaller change I can measure than a big one I can’t.
On call is where you learn what the system really does.
Areas to probe
Worth a question in your own interview
Asked, with less to show
Works closely with product and design: joined once designs were final.
Interview transcript
6 questions, in order
Show transcript
Prepared from the interview recording and the role’s requirements.
Powered by Hireanex · Anyone with this link can view this report · Link valid until October 19, 2026
You choose what to send: the link expires, and Stop sharing works at once. The client’s answer comes back to your team only.
Request a demo