63 questions. Open any one for the answer.
A $483,589 answer Describe an MCP server, AI agent integration, or LLM tool-calling system you personally built. What APIs did it expose, how did you enforce what the agent was allowed to do, and how did you log agent activity?
A $483,589 answer Describe a production API you built that handled untrusted or automated callers. How did you handle authentication, rate limiting, input validation and auditing?
A $288,651 answer Describe a production system you built with Claude Code, skills, MCP, subagents, hooks and/or gates. What made it reliable, and what broke along the way?
A $1,144,267 answer Walk me through a real search or recommendation system you built, not a CRUD app. What was the ranking logic, and what broke first at scale?
A $665,013 answered no answer Name one production system where 5+ coding agents opened real PRs against one large codebase with Linear as the source of intent. What did agents write back to issues, what were they forbidden to write, and why? Numbers, not principles.
A Mexico FDE answer Describe a Python backend service you built that had to run unattended: how did you handle queues, retries, idempotency and failure alerts?
A answer Describe a scraper you built that ran in production. What broke it, and how did you find out it had broken?
A $1,101,306 answer Describe one production n8n workflow you built, including integrations and how you validated its reliability.
A $935,596 answer Tell us about 2 to 3 AI or automation solutions you've personally built. For each: the problem, the stack, what you built, and the measurable result.
A $494,665 answer Describe the most complex voice AI agent you have shipped to production. What was the workflow, how did you handle interruptions and confirmations, and what broke in the real world?
A Upwork itself, avg bid $67.92, max $350 answered no answer Describe one GRC automation you personally built and shipped. What manual process did it replace, what was the stack, roughly how much time did it save, and who maintained it after you left? Include what broke or didn't work as expected.
A $325,374 answered no answer Have you fine-tuned a coding LLM before? Send benchmark/results.
A $1,115,131 answer Describe a real-time voice agent you've personally built or tuned. What was your end-to-end latency, and what got it there?
A $1,827,106 answer Describe an AI/ML application you personally built. What did you design and code yourself?
B $288,651 answer Walk through a specific agent/LLM behaviour bug you diagnosed from logs or traces: the symptom, the real root cause, and how you proved the fix improved things without breaking others.
B $373,434 answer A short 5 to 10 line story of a bug you fixed using Claude Code, including one thing the agent got wrong and how you caught it.
B $274,389 answer Tell us about a tool or automation you built that broke after launch. What caused it, how did you fix it, and what did you change to reduce the chance of it happening again?
B answer Explain how you would detect your own scraper silently breaking." with the client adding: *the second half is the part we care about most*
B answer Given a 60-case fixture set at 48% pass, how do you avoid overfitting fixes to the fixtures?
B answer Where would you put the check that prevents a false 'I moved it' claim, and why there rather than in the prompt?
B $665,013 answer Delegation silently did nothing while the webhook endpoint looked healthy. What's your first suspect, how do you confirm it inside an hour, and what monitoring stops it recurring?
B $665,013 answer Linear redelivers webhooks and agents retry. A duplicate session event double-dispatched a task and produced two PRs. Where does idempotency live, Linear, the webhook layer, or your orchestrator?
B $318,518 answer Tell us about a technical decision you made in the last two years that you later regretted. What did you do differently afterwards?
B answer Give an example of an AI agent or LLM vulnerability you found and how you reproduced it.
B $494,665 answered no answer Describe a voice interaction that failed in production: noise, an accent, silence, or an action that did not commit, and how you handled it.
B $1,144,267 answer If an ecommerce store's revenue suddenly dropped 30% while ad spend stayed the same, how would you diagnose the cause before making changes?
C $309,310 answer You have an 8-hour chest-mounted body-camera video and need to detect 20 behaviours, some lasting 10 seconds and some 30+ minutes. How would you architect it? Would you send the entire 8-hour video through Gemini in one prompt? Why or why not?
C $309,310 answer How would you measure false positives, false negatives, precision, recall and overall performance of a video LLM system?
C $309,310 answer When would you use prompting vs RAG vs fine-tuning for this problem?
C $1,144,267 answer Our catalog has one physical part mapped to 15+ vehicles via a fitment table. Design, out loud, a filter system that stays fast as that table grows into the millions of rows.
C $90,310 answer Walk through how you'd design accept/reject decisioning that must respond under 200ms at 50 requests a second.
C $318,518 answer A student has spent 100 hours talking to our life coach over six months. How would you architect the memory?
C $665,013 answer A conductor splits an epic into ~20 dependency-ordered work packages built by parallel agents. Model it in Linear. What breaks at 50 agent-driven updates an hour?
C $1,115,131 answer How would you approach barge-in, the user starting to talk while the bot is speaking, in a browser where echo cancellation can't see the TTS audio?
C $325,374 answer How would you create 10,000 executable repo-level coding tasks?" / "How would you prevent SWE-bench data contamination?" / "SFT, DPO or RL first, and why?
C $44,970 answer How would you prevent Claude from accidentally deploying changes to the wrong client's live store?" / "How would you securely manage credentials for many Shopify stores?
C $288,651 answer How do you decide which model and reasoning settings to use for a given agent task, and how do you design gates to stop an agent doing the wrong thing?
C $84,145 answer How would you prevent the AI from presenting unsupported assumptions as verified research findings?" / "How would you retain sources so every important conclusion can be independently verified?
C Upwork GRC answer How do you make AI-generated compliance output defensible to an auditor?
D $1,144,267 answer Share 3 to 5 specific rules from a CLAUDE.md you actively use, with the reasoning behind each. What project-specific issue or Claude behaviour prompted you to add them?
D $1,144,267 answer Claude Code proposes a 12-file change to fix a slow filter query. Before you approve it, what do you personally check, and what would make you reject or shrink the diff?
D $1,144,267 answer You personally have Claude Max right now. What tier, and what's the last thing Claude Code did that surprised you this week, good or bad?
D $1,144,267 answer Do you use Claude Code daily? You will be told in an interview to share your screen, and the interviewer is a Claude Code specialist.
D German SaaS, 97,351 hours answer Which AI coding tools do you use daily, and one specific trick or setup that makes you faster than the average user.
D $373,434 answer You use Claude Code daily to ship real code, not just chat. Explain your setup: CLAUDE.md, subagents, plan mode, hooks, how you review what the agent produced.
D $339,494 answer How has AI changed the way you do data work in the last year? Be specific: which tools, what you've built, and what you stopped doing by hand.
D $275,137 answer How comfortable are you with a Python debugger? With PyTest?
E answer Do you have experience designing reproducible test suites or benchmarks?
E $226,195 answer Describe your approach to testing and improving QA
E $373,434 answer Your pre-deploy checklist for a production release, as a bullet list.
E $288,651 answer Measure before and after with evidence, not vibes.
E $325,374 answer Success will be measured using actual test-passing repository tasks, not training loss.
E $1,115,131 answered no answer prove them with before/after measurements: latency numbers, and recorded test conversations.
F $483,589 answer Published or open-source MCP servers. Include links in your application.
F $988,860 answer Experience building MCP servers and tools, a concrete, checkable skill, not just 'uses AI tools'
F $609,912 answered no answer Contributions to anthropic-sdk-python or claude-agent-sdk-python/typescript, or filed reproducible bug reports against them.
F $609,912 answer any open-source contributions or repos we can review
F a literal screening field on at least 4 jobs answer Include a link to your GitHub profile and/or website
F German SaaS answer A link or screenshots of something real you shipped: repo, staging URL, screen recording, anything we can actually look at.
F Mexico FDE answer Please include one link (repo, demo or write-up) that shows a messaging agent or backend service you built and operated.
F Songsterr, avg bid $102.15 answer Start your proposal with the strongest thing you have personally shipped to production, with measurable results and your specific role.
F $376,190 answered no answer three examples of AI systems you personally built and a short video walking us through your work
F $935,596 answer a portfolio of real work you can walk us through