Customer Success Engineer (LatAm)

What you would do

• Reproduce and investigate user-reported issues end-to-end, identifying root causes in code, configuration, or AI model behavior
• Debug across the web and AI stack using logs, telemetry, network inspection, and database queries
• Distinguish between legitimate bugs, model edge cases, UX friction, and user configuration errors
• Triage issues with sound judgment: resolve those fixable via configuration or guidance, escalate true engineering defects
• Document resolution approaches and systemic issues, creating runbooks that reduce future repeat incidents

Who they want

• Computer Science or Software Engineering degree from top-tier school, or 2-5 years supporting web applications in growth-stage startups
• Experience building or deeply using modern web frameworks such as React, Node, Flask, or Next.js
• Familiarity with AI systems: experience playing with LLMs, agents, or generative models in real applications
• Capability to interpret model outputs, recognize failure modes and hallucinations, and understand how feedback impacts performance
• Comfortable debugging infrastructure: reading logs and network traffic, running database queries, using monitoring and telemetry tools

Main skills

What the interview asks about

1. 1.Bug reproduction and root cause

   Finding the real cause of a reported issue requires systematic investigation and technical intuition; this shows whether you can think through the stack methodically.

   For example: “A user says they completed their application but it shows 'incomplete' in their dashboard. Where would you look first to debug this: frontend, backend, database, or AI system? Walk through your approach.”

2. 2.AI system behavior analysis

   Understanding when AI model behavior is working as designed versus actually broken is difficult; this assesses your capability with AI debugging.

   For example: “A user disputes their interview score, claiming it's unfair. The output looks correct but feels wrong. Is this a hallucination, prompt issue, data problem, or expectation mismatch? How would you investigate?”

3. 3.Triage and escalation judgment

   Knowing which issues need engineering versus which you can resolve yourself prevents bottlenecks and shows good judgment; this tests accountability and decision-making.

   For example: “You get three reports in an evening: one user can't log in, one user sees a typo in AI feedback, and one user's model scores seem inconsistent. Prioritize these, decide which get escalated and which you handle, and explain your reasoning.”

4. 4.Technical communication

   Users and engineers need clear, precise updates; this evaluates whether you can explain technical problems in accessible language.

   For example: “After debugging a model scoring issue, write a 100-word explanation for both a user and an engineer explaining what went wrong, what you found, and what was fixed.”

5. 5.Systemic pattern recognition

   Beyond individual bugs, spotting patterns that signal larger risks shows strategic thinking; this matters for platform health and prevents repeated failures.

   For example: “Over a week, you see 8 support tickets about model scores seeming inconsistent with user input. Each is slightly different, but you suspect a common underlying issue. How would you investigate and escalate this to product leadership?”

A task you may get

Review 2-3 sample support tickets. For each, outline your investigation approach, tools you'd use, where in the stack you'd look, and whether you'd resolve or escalate. Provide reasoning for each decision.

How to prepare

• Refresh your knowledge of modern web application debugging: browser developer tools, network inspection, common frontend and backend failure patterns
• Study how LLM APIs and agent frameworks work; understand common failure modes like hallucinations, prompt sensitivity, and feedback loops
• Review your experience supporting SaaS products: what issues did you encounter, how did you diagnose them, what made triage decisions clear or ambiguous
• Prepare examples of bugs you've investigated: how you found them, what surprised you, how you communicated findings to technical teams

The facts

Pay

$35k–50k/yr

Hours

Full time

Where

Remote

Open to

MEX, CRI, SLV, GTM, PAN, ARG, BOL, BRA, CHL, COL, ECU, GUY, PRY, PER, URY

Field

Software Engineering

Posted

5/5/2026

Places left

1

We wrote this page from the public Mercor listing. It may be out of date, so read the full posting before you apply.