Senior Product Manager, Conversational AI Chatbot & Agent Quality
Singapore, Singapore·Posted 1mo ago
web3blockchainllm
<div class="ace-line ace-line old-record-id-doxuseysYUio6Qia64JLLAwE7dh"> <div data-page-id="doxusokjWsaOkSCIjzixAfRM3sd" data-docx-has-block-data="false"> <div class="ace-line ace-line old-record-id-doxusaUYeCmu82WSkkm5KDd00db"> <div data-page-id="JAF7dFJcWoUusRx7RKkuM14BsYc" data-docx-has-block-data="false"> <div class="ace-line ace-line old-record-id-OqGZdUPi6oJlXzxlmNiu3gAmsVg"> <div data-page-id="V6OedkjJzouK9jxNIfuuhjt9sDc" data-docx-has-block-data="false"> <div data-page-id="AEW3d0Y2noLuIcxROuFubTLpsZd" data-docx-has-block-data="false"> <div class="ace-line ace-line old-record-id-doxuseysYUio6Qia64JLLAwE7dh"> <div data-page-id="doxusokjWsaOkSCIjzixAfRM3sd" data-docx-has-block-data="false"> <div class="ace-line ace-line old-record-id-doxusaUYeCmu82WSkkm5KDd00db"> <div data-page-id="AEW3d0Y2noLuIcxROuFubTLpsZd" data-docx-has-block-data="false"> <div data-page-id="JAF7dFJcWoUusRx7RKkuM14BsYc" data-docx-has-block-data="false"> <div class="ace-line ace-line old-record-id-OqGZdUPi6oJlXzxlmNiu3gAmsVg"><em>OKX will be prioritising applicants who have a current right to work in Singapore, and do not require OKX's sponsorship of a visa.<br><br></em></div> </div> <h2 class="heading-2 ace-line old-record-id-doxuslsyQOGHoiYb47TiA1n51Th"><strong>Who We Are</strong></h2> <div class="ace-line ace-line old-record-id-doxustOiFVk8C3Uy4rotl5Nem5f"> <div data-page-id="PNNZdiw4Yo1ZOmx8btbucw8qsLG" data-docx-has-block-data="false"> <div class="ace-line ace-line old-record-id-Q479dQIlZozY6cxcPNFuoY1rsxe"> <div class="rich-text-paragraph" data-eleid="27"> <div class="ace-line ace-line old-record-id-RKOAdw3kVoh5EQxcr2juP3i0sTb"> <div class="ace-line ace-line old-record-id-Cfb8dvi9voxFkWxhNcmuJX50sZb">At OKX, we believe that the future will be reshaped by crypto, and ultimately contribute to every individual's freedom.</div> <div class="ace-line ace-line old-record-id-Cfb8dvi9voxFkWxhNcmuJX50sZb"> </div> <div class="ace-line ace-line old-record-id-Cfb8dvi9voxFkWxhNcmuJX50sZb">OKX is a leading crypto exchange, and the developer of OKX Wallet, giving millions access to crypto trading and decentralized crypto applications (dApps). OKX is also a trusted brand by hundreds of large institutions seeking access to crypto markets. We are safe and reliable, backed by our Proof of Reserves. </div> <div class="ace-line ace-line old-record-id-Cfb8dvi9voxFkWxhNcmuJX50sZb"> </div> <div class="ace-line ace-line old-record-id-Cfb8dvi9voxFkWxhNcmuJX50sZb">Across our multiple offices globally, we are united by our core principles: <em>We Before Me</em>, <em>Do the Right Thing</em>, and <em>Get Things Done</em>. These shared values drive our culture, shape our processes, and foster a friendly, rewarding, and diverse environment for every OK-er.<br><br> <div data-page-id="Kpucdjv7JoAcSZxSf7PuRl5Yscb" data-lark-html-role="root" data-docx-has-block-data="false"> <div class=" old-record-id-Cfb8dvi9voxFkWxhNcmuJX50sZb">OKX is part of OKG, a group that brings the value of Blockchain to users around the world, through our leading products OKX, OKX Wallet, OKLink and more.</div> </div> </div> </div> </div> </div> </div> </div> <div class="ace-line ace-line old-record-id-doxusq2WnfR822THsuqUosdSzFu"> </div> <h2 class="heading-2 ace-line old-record-id-doxus9qcafz8J9vhi3nwZgrckWg"><strong>About The Opportunity</strong></h2> <div data-lark-html-role="root"> <div class="rich-text-paragraph" data-eleid="15"> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">We are looking for an execution-focused Product Manager who has built and improved conversational AI products in production — and has business results to prove it. A strong plus is hands-on experience with agent evaluation harnesses or internal agent platform product design: you've defined the systems that test, score, and operate agents at scale, not just shipped the agents themselves.</p> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">You work in logs and specs, not just decks. You know what a bad retrieval chunk looks like, you've personally written labeling guidelines, and you can point to a quarter where your work moved resolution rate by double digits.</p> </div> </div> <h2 class="font-claude-response-body break-words whitespace-normal leading-[1.7]"><strong>What We Are Looking For</strong></h2> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]">You have hands-on experience building and operating conversational AI products in production — not just shipping agents, but owning the quality systems, data pipelines, and operational platforms that keep them reliable at scale. Ideal candidates will have background in one or more of the following areas:</p> <ul class="[li_&]:mb-0 [li_&]:mt-1 [li_&]:gap-1 [&:not(:last-child)_ul]:pb-1 [&:not(:last-child)_ol]:pb-1 list-disc flex flex-col gap-1 pl-8 mb-3"> <li class="font-claude-response-body whitespace-normal break-words pl-2"><strong>Knowledge Base & Data Quality</strong> — knowledge base architecture, retrieval quality tuning, content governance, labeling pipelines, annotation guidelines, training data impact tracking, and dataset freshness management</li> <li class="font-claude-response-body whitespace-normal break-words pl-2"><strong>Agent Evaluation & Quality Assurance</strong> — evaluation harness design, test case schemas, automated scoring rubrics (correctness, groundedness, tool-use accuracy), LLM-as-judge evaluation, regression testing for non-deterministic systems, and feedback-driven improvement loops</li> <li class="font-claude-response-body whitespace-normal break-words pl-2"><strong>Chatbot Operations & Dialogue Design</strong> — SOP-to-agent-flow translation, edge case handling, escalation path design, log-based failure triage, and metrics ownership (resolution rate, fallback rate, per-intent accuracy, CSAT)</li> <li class="font-claude-response-body whitespace-normal break-words pl-2"><strong>Agent Runtime & Observability Platforms</strong> — agent runtime product requirements, tool permission models, task configuration interfaces, developer-facing observability dashboards, failure alerting logic, and debugging workflows</li> <li class="font-claude-response-body whitespace-normal break-words pl-2"><strong>Human-in-the-Loop Workflows</strong> — low-confidence case routing, reviewer task interface design, correction data capture, and feedback loop integration back into training or knowledge pipelines</li> </ul> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]"><strong>Chatbot & Knowledge Base (Core)</strong></p> <ul class="[li_&]:mb-0 [li_&]:mt-1 [li_&]:gap-1 [&:not(:last-child)_ul]:pb-1 [&:not(:last-child)_ol]:pb-1 list-disc flex flex-col gap-1 pl-8 mb-3"> <li class="font-claude-response-body whitespace-normal break-words pl-2">Built or rebuilt a knowledge base — defined structure, wrote/reviewed content, fixed retrieval quality, saw metrics improve</li> <li class="font-claude-response-body whitespace-normal break-words pl-2">Designed SOPs that became agent flows — mapped real business processes, handled edge cases, shipped as working dialogue flows</li> <li class="font-claude-response-body whitespace-normal break-words pl-2">Owned a labeling pipeline — wrote annotation guidelines, QA'd batches, tracked whether labeled data moved production metrics</li> <li class="font-claude-response-body whitespace-normal break-words pl-2">Moved a metric that mattered — resolution rate, fallback rate, CSAT — and can explain exactly what changed</li> </ul> <p class="font-claude-response-body break-words whitespace-normal leading-[1.7]"><strong>Agent Harness & Platform Product (Strong Plus)</strong></p> <ul class="[li_&]:mb-0 [li_&]:mt-1 [li_&]:gap-1 [&:not(:last-child)_ul]:pb-1 [&:not(:last-child)_ol]:pb-1 list-disc flex flex-col gap-1 pl-8 mb-3"> <li class="font-claude-response-body whitespace-normal break-words pl-2">Designed an agent evaluation harness: defined test case schemas, scoring rubrics, and spec'd automated evaluation pipelines with engineering</li> <li class="font-claude-response-body whitespace-normal break-words pl-2">Product-designed an internal agent platform: defined requirements for agent runtime — tool permission models, task configuration interfaces, developer-facing observability dashboards, and failure debugging workflows; owned the roadmap and shipped iteratively</li> <li class="font-claude-response-body whitespace-normal break-words pl-2">Closed the eval-to-improvement loop: used harness output to prioritize knowledge fixes, prompt revisions, or flow changes — not just reported scores but drove action from them</li> <li class="font-claude-response-body whitespace-normal break-words pl-2">Designed human-in-the-loop review workflows: low-confidence case routing, reviewer task interfaces, correction data capture and feedback loop back into training or knowledge pipelines</li> </ul> <h2 class="heading-2 ace-line old-record-id-doxushHxPgvpIV0pJrijghkWDWe"><strong>What You’ll Be Doing </strong></h2> <div data-lark-html-role="root"> <div class="rich-text-paragraph" data-eleid="70"><strong><span class="text-only" data-eleid="71"><span class="text-only">Chatbot Operations</span></span></strong></div> <ul class="richTextDocs-unOrderList richTextDocs-unOrderList-disc" data-eleid="73"> <li class="richTextDocs-listItem" data-eleid="74"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="75"><span class="text-only">Knowledge base ownership: structure, content quality, retrieval coverage, freshness governance</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="78"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="79"><span class="text-only">SOP & dialogue flow design: business process → agent flow → edge case handling → escalation paths</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="82"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="83"><span class="text-only">Labeling pipeline: annotation specs, annotator QA, training batch impact tracking</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="86"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="87"><span class="text-only">Daily quality work: log review, failure triage, weekly knowledge/flow update cadence</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="90"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="91"><span class="text-only">Metrics ownership: resolution rate, fallback rate, per-intent accuracy, CSAT</span></span></div> </li> </ul> <div class="rich-text-paragraph" data-eleid="93"><strong><span class="text-only" data-eleid="94"><span class="text-only">Agent Harness & Platform Product</span></span></strong></div> <ul class="richTextDocs-unOrderList richTextDocs-unOrderList-disc" data-eleid="96"> <li class="richTextDocs-listItem" data-eleid="97"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="98"><span class="text-only">Define and maintain </span></span><strong><span class="text-only" data-eleid="99"><span class="text-only">agent evaluation frameworks</span></span></strong><span class="text-only" data-eleid="100"><span class="text-only">: test case design, automated scoring criteria, regression test coverage</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="103"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="104"><span class="text-only">Own the </span></span><strong><span class="text-only" data-eleid="105"><span class="text-only">quality feedback loop</span></span></strong><span class="text-only" data-eleid="106"><span class="text-only">: harness results → prioritized fixes → re-evaluation → </span><span class="text-only text-with-abbreviation text-with-abbreviation-bottomline">production deployment</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="109"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="110"><span class="text-only">Partner with engineering to </span></span><strong><span class="text-only" data-eleid="111"><span class="text-only">define product requirements for agent runtime</span></span></strong><span class="text-only" data-eleid="112"><span class="text-only">: spec observability features, tool call monitoring interfaces, failure alerting logic, and developer-facing debugging tools — own the backlog, not the </span><span class="text-only text-with-abbreviation text-with-abbreviation-bottomline">ops</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="115"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="116"><span class="text-only">Design </span></span><strong><span class="text-only" data-eleid="117"><span class="text-only">human-in-the-loop workflows</span></span></strong><span class="text-only" data-eleid="118"><span class="text-only">: case routing logic, reviewer interfaces, correction data capture</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="121"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="122"><span class="text-only">Track agent version performance over time; maintain eval dashboards that teams actually use</span></span></div> </li> </ul> </div> <h2 class="heading-2 ace-line old-record-id-doxusWnZPeJsdMU53QGew90VQeh"><strong>What We Look For In You </strong></h2> <div data-lark-html-role="root"> <ul class="richTextDocs-unOrderList richTextDocs-unOrderList-disc" data-eleid="127"> <li class="richTextDocs-listItem" data-eleid="128"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="129"><span class="text-only">3–6 years PM experience; </span></span><strong><span class="text-only" data-eleid="130"><span class="text-only">minimum 2 years as primary owner of a production chatbot or AI agent product</span></span></strong></div> </li> <li class="richTextDocs-listItem" data-eleid="133"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="134"><span class="text-only">Quantified business results: can describe baseline metrics, what you did, and outcome in numbers</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="137"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="138"><span class="text-only">Hands-on knowledge base, labeling, and conversation analysis experience (not just oversight)</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="141"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="142"><span class="text-only">Familiar with at least one chatbot/agent platform (Coze, Dify, Dialogflow, or similar)</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="145"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="146"><span class="text-only">Mandarin Chinese fluency required; English proficiency required</span></span></div> </li> </ul> </div> <h2 class="heading-2 ace-line old-record-id-doxusWnZPeJsdMU53QGew90VQeh"><strong>Nice-To-Haves </strong></h2> <div data-lark-html-role="root"> <ul class="richTextDocs-unOrderList richTextDocs-unOrderList-disc" data-eleid="151"> <li class="richTextDocs-listItem" data-eleid="152"> <div class="richTextDocs-listItem__text"><strong><span class="text-only" data-eleid="153"><span class="text-only">Designed an agent eval harness</span></span></strong><span class="text-only" data-eleid="154"><span class="text-only">: written test case specs, defined scoring rubrics (correctness, groundedness, tool-use accuracy), and spec'd the automated evaluation pipeline with engineering</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="157"> <div class="richTextDocs-listItem__text"><strong><span class="text-only" data-eleid="158"><span class="text-only">Product-designed an internal agent platform</span></span></strong><span class="text-only" data-eleid="159"><span class="text-only">: defined product requirements for agent runtime — tool permission models, task configuration interfaces, developer-facing observability and debugging workflows; owned roadmap and shipped iteratively</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="162"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="163"><span class="text-only">Experience with </span></span><strong><span class="text-only" data-eleid="164"><span class="text-only">LLM-as-judge evaluation</span></span></strong><span class="text-only" data-eleid="165"><span class="text-only">: has used model-based scoring in a harness and understands its blind spots</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="168"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="169"><span class="text-only">Familiar with agent observability tooling (LangSmith, Langfuse, or internal equivalents) — to define what the product needs to surface, not to operate them</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="172"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="173"><span class="text-only">Experience </span></span><strong><span class="text-only" data-eleid="174"><span class="text-only">spec'ing regression testing for non-deterministic systems</span></span></strong><span class="text-only" data-eleid="175"><span class="text-only">: knows how to define quality regression detection when </span><span class="text-only text-with-abbreviation text-with-abbreviation-bottomline">LLM</span><span class="text-only"> outputs vary</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="178"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="179"><span class="text-only">Has written </span></span><strong><span class="text-only" data-eleid="180"><span class="text-only">product specs for human-in-the-loop workflows</span></span></strong><span class="text-only" data-eleid="181"><span class="text-only">: low-confidence case routing, reviewer task interfaces, correction data capture and feedback loop design</span></span></div> </li> <li class="richTextDocs-listItem" data-eleid="184"> <div class="richTextDocs-listItem__text"><span class="text-only" data-eleid="185"><span class="text-only">Background in customer service, operations, or financial services domain</span></span></div> </li> </ul> </div> <h2><strong>Perks & Benefits </strong></h2> <ul class="list-bullet1"> <li class="ace-line ace-line old-record-id-Gb04dtjtMoHmldxfAjXuI6P4snC" data-list="bullet"> <div>Competitive total compensation package</div> </li> <li class="ace-line ace-line old-record-id-Gb04dtjtMoHmldxfAjXuI6P4snC" data-list="bullet">L&D programs and education subsidy for employees' growth and development</li> <li class="ace-line ace-line old-record-id-doxusfCPfNQIPMLDddcFHLcCJRC" data-list="bullet"> <div>Various team building programs and company events</div> </li> <li class="ace-line ace-line old-record-id-doxusfCPfNQIPMLDddcFHLcCJRC" data-list="bullet">Wellness and meal allowances</li> <li class="ace-line ace-line old-record-id-doxusfCPfNQIPMLDddcFHLcCJRC" data-list="bullet">Comprehensive healthcare schemes for employees and dependants</li> <li class="ace-line ace-line old-record-id-doxusfCPfNQIPMLDddcFHLcCJRC" data-list="bullet">More that we love to tell you along the process!</li> </ul> <p> </p> <p><span style="color: rgb(255, 255, 255);">#LI-WWW</span><br><span style="color: rgb(255, 255, 255);">#LI-ONSITE</span></p> </div> </div> </div> </div> </div> </div> </div> </div> </div> </div> </div><div class="content-conclusion"><div data-lark-html-role="root"><span class="text-only" data-eleid="18"><span class="text-only"><span class="text-only" data-eleid="6">Notice:<br></span></span></span> <div data-lark-html-role="root"><span class="text-only" data-eleid="26"><span class="text-only">All official </span><span class="text-only text-with-abbreviation text-with-abbreviation-bottomline">OKX</span><span class="text-only"> vacancies are published on this website.</span></span> <span class="text-only" data-eleid="28"><span class="text-only">While roles may appear on selected third-party platforms from time to time, information on other sites may be inaccurate or outdated. </span></span><strong><span class="text-only" data-eleid="29"><span class="text-only">If in doubt, please apply directly through our official careers website.</span></span></strong></div> </div> <div data-lark-html-role="root"><span class="text-only" data-eleid="18"><span class="text-only">Information collected and processed as part of the recruitment process of any job application you choose to submit is subject to </span><span class="text-only text-with-abbreviation text-with-abbreviation-bottomline">OKX</span><span class="text-only">'s </span></span><a class="link rich-text-anchor __anchor-intercept-flag__ text-content-link" href="https://www.okx.com/en-eu/help/okx-candidate-privacy-notice" target="_blank" data-eleid="19" data-lark-is-custom="true" data-lark-link="true">Candidate Privacy Notice</a><span class="text-only" data-eleid="20"><span class="text-only">.</span></span></div></div>