Director, Applied AI

We need to output HTML fragment with sections: About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule. Only include sections that have real specific content. Must not repeat job title. Must not include boilerplate like EEO, legal, etc. Must preserve all substantive info. We need to parse the posting. Let's extract sections: - Intro paragraph: "ZoomInfo is where careers accelerate..." that's about the role? Could be "About the role". We'll treat as About the role. - Then description of role: "You'll lead the team that builds the intelligence every ZoomInfo AI agent reasons over..." etc. That's responsibilities. - Then bullet list under "What You'll Do": multiple sentences each starting with "You will". Those are responsibilities. - Then "Must-Have What You Bring": list of requirements/qualifications. - Then "Preferred": list of preferred qualifications. - Then compensation: "Actual compensation offered will be based on factors..." plus salary range: "$233,100—$366,300 USD". Also mention bonus, commission, equity. - Benefits: "We want our employees and their families to thrive. In addition to comprehensive benefits we offer holistic mind, body and lifestyle programs designed for overall well-being. Learn more about ZoomInfo benefits here." - Also maybe "Schedule": not mentioned. No explicit schedule info. So we should not include Schedule section. - Pay: we have salary range and mention additional compensation. So we can include Pay section. - Benefits: we have description of benefits. - Skills: maybe we can extract from Must-Have and Preferred as skills. But we already have Requirements/Qualifications sections. The instruction: use headings for section headings (About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule). We need to decide which to use. Could combine Requirements and Qualifications? But they are separate headings allowed. We need to include sections only when posting has real specific content. We have content for Responsibilities, Requirements (Must-Have), Qualifications (Preferred), Benefits, Pay. Also About the role (intro). Skills maybe not separate; we could embed skills within Requirements/Qualifications. But we could also create Skills section summarizing technical skills: classical ML, LLM, agentic systems, SQL, etc. However we must not invent or summarize away detail; we can list them as they appear. The posting includes: "deep classical machine learning and data science expertise (supervised learning, feature engineering, statistical inference, experiment design, strong SQL) alongside production LLM and agentic systems". Also "owned inference cost, latency, and capacity". Also "experience in propensity modeling, ranking and retrieval, clustering, or entity resolution at scale". Also "worked on web-scale language processing over multilingual, noisy text, knowledge graphs, or user memory for agents". Also "experience with post-training and distillation, open-weight model serving, or AI governance and safety practices". Those are specific skills. So we could create a Skills section with those items. But we must not repeat content; we can still include them in Skills and also in Requirements/Qualifications? Might be duplication. Better to follow typical structure: About the role (intro), Responsibilities (What You'll Do), Requirements (Must-Have), Qualifications (Preferred), Skills (maybe combine technical skills from both), Benefits, Pay. Ensure we don't create empty sections. Let's extract each piece. About the role: The first paragraph(s) before "What You'll Do". Let's capture: "ZoomInfo is where careers accelerate. We move fast, think boldly, and empower you to do the best work of your life. You’ll be surrounded by teammates who care deeply, challenge each other, and celebrate wins. With tools that amplify your impact and a culture that backs your ambition, you won’t just contribute. You’ll make things happen–fast. You'll lead the team that builds the intelligence every ZoomInfo AI agent reasons over — what's true about companies and the people in them, how they relate, and what they're buying. This role owns the B2B data graph strategy end to end, from training data through model serving, blending classical machine learning and data science with LLM and agentic systems. You'll set the technical bar for a small, senior team while staying hands-on, shipping code and prototypes alongside the engineers you lead." That seems appropriate for About the role. Responsibilities: The bullet list under "What You'll Do". Each sentence starts with "You will". We'll capture them as list items. List: - extend ZoomInfo's B2B data graph into the long tail, making it the most comprehensive and accurate resource available, including companies with little public footprint. - lead the team's work on agent memory, distinguishing what a user has supplied from what a system of record already owns. - choose the right method for each problem — classical machine learning, language models, or code — based on measured evidence, and stop work that won't pay off. - define the evaluation bar for the team's models and agents, building evaluation datasets, regression gates, and experiment designs that make quality claims trustworthy. - own inference cost, latency, and capacity as core disciplines alongside model quality, including build-versus-buy and distillation decisions. - hire and develop machine learning engineers, data scientists, and research engineers, growing senior engineers into technical leaders. - represent the team's work across product, platform, security, and legal, and present results and their limits clearly to executives. - set the standard for how the team uses agentic coding tools, pairing precise specifications with rigorous code review. We'll put each as

. Requirements (Must-Have): The list under "Must-Have What You Bring". Let's capture each bullet. - You have significant, demonstrated experience building and shipping production machine learning systems, and you lead by building alongside your team — capability matters more than years. - You have a track record of hiring and developing senior machine learning engineers and data scientists against a high bar. - You stay hands-on today: shipping code, prototyping independently, and using agentic coding tools daily with rigorous review. - You bring deep classical machine learning and data science expertise (supervised learning, feature engineering, statistical inference, experiment design, strong SQL) alongside production LLM and agentic systems, with the judgment to choose between them, including setting evaluation standards such as leakage-safe validation and calibration. - You have owned inference cost, latency, and capacity alongside model quality, including build-versus-buy decisions, and you communicate results and their limits clearly to executive stakeholders. We'll put as list. Qualifications (Preferred): The list under "Preferred". Capture: - You bring entrepreneurial experience — founding a company, or taking a product from inception to paying customers as a founding or early engineer. - You have experience in propensity modeling, ranking and retrieval, clustering, or entity resolution at scale. - You have worked on web-scale language processing over multilingual, noisy text, knowledge graphs, or user memory for agents. - You have experience with post-training and distillation, open-weight model serving, or AI governance and safety practices (ISO/IEC 42001, NIST AI RMF). We'll put as list. Skills: Could extract technical skills from both Must-Have and Preferred, but to avoid duplication we could create a Skills section that lists specific technical competencies: classical ML (supervised learning, feature engineering, statistical inference, experiment design, strong SQL), production LLM and agentic systems, evaluation standards (leakage-safe validation, calibration), inference cost, latency, capacity, build-versus-buy, distillation, agentic coding tools, propensity modeling, ranking and retrieval, clustering, entity resolution, web-scale language processing, multilingual noisy text, knowledge graphs, user memory for agents, post-training, distillation, open-weight model serving, AI governance and safety practices (ISO/IEC 42001, NIST AI RMF). However many of these are already in Requirements/Qualifications. The instruction says "Preserve ALL substantive information; do NOT invent or summarize away detail." If we create Skills section that repeats content, we are not inventing but we are duplicating. Is duplication allowed? Not explicitly prohibited, but better to avoid redundancy. However we need to include a Skills section only if there is real specific content for it. The posting does have specific skill mentions; we could treat them as Skills section and not duplicate them elsewhere. But then we would need to move those details from Requirements/Qualifications to Skills, which would remove them from those sections, potentially losing some info. But we could keep Requirements/Qualifications as more about experience and track record, and Skills as technical abilities. That seems reasonable. Let's define: Requirements (Must-Have): focus on experience, leadership, hands-on, track record, etc. Could keep the bullet points but maybe rephrase to focus on experience rather than technical skills. However the Must-Have includes technical expertise bullet. That could be moved to Skills. But we need to preserve all substantive info; moving is okay as long as we keep it somewhere. Thus we can split: Requirements (Must-Have) bullets: - You have significant, demonstrated experience building and shipping production machine learning systems, and you lead by building alongside your team — capability matters more than years. - You have a track record of hiring and developing senior machine learning engineers and data scientists against a high bar. - You stay hands-on today: shipping code, prototyping independently, and using agentic coding tools daily with rigorous review. - You have owned inference cost, latency, and capacity alongside model quality, including build-versus-buy decisions, and you communicate results and their limits clearly to executive stakeholders. (We omitted the deep classical ML expertise bullet; that'll go to Skills.) Qualifications (Preferred) bullets: - You bring entrepreneurial experience — founding a company, or taking a product from inception to paying customers as a founding or early engineer. - You have experience in propensity modeling, ranking and retrieval, clustering, or entity resolution at scale. - You have worked on web-scale language processing over multilingual, noisy text, knowledge graphs, or user memory for agents. - You have experience with post-training and distillation, open-weight model serving, or AI governance and safety practices (ISO/IEC 42001, NIST AI RMF). Skills: list technical competencies: - Deep classical machine learning and data science expertise (supervised learning, feature engineering, statistical inference, experiment design, strong SQL) - Production LLM and agentic systems - Judgment to choose between classical ML, language models, or code based on measured evidence - Setting evaluation standards such as leakage-safe validation and calibration - Inference cost, latency, and capacity optimization - Build-versus-buy and distillation decisions - Agentic coding tools usage with rigorous review - Propensity modeling, ranking and retrieval, clustering, entity resolution at scale - Web-scale language processing over multilingual, noisy text - Knowledge graphs - User memory for agents - Post-training and distillation - Open-weight model serving - AI governance and safety practices (ISO/IEC 42001, NIST AI RMF) We need to ensure we don't lose any info: The original Must-Have also included "You bring deep classical machine learning and data science expertise (supervised learning, feature engineering, statistical inference, experiment design, strong SQL) alongside production LLM and agentic systems, with the judgment to choose between them, including setting evaluation standards such as leakage-safe validation and calibration." That's covered in Skills bullets. Also "You have owned inference cost, latency, and capacity alongside model quality, including build-versus-buy decisions, and you communicate results and their limits clearly to executive stakeholders." The ownership part is in Requirements; the communication part also in Requirements. Good. Now Benefits: we have paragraph: "We want our employees and their families to thrive. In addition to comprehensive benefits we offer holistic mind, body and lifestyle programs designed for overall well-being. Learn more about ZoomInfo benefits here." We'll put as

. Pay: we have: "Actual compensation offered will be based on factors such as the candidate’s work location, qualifications, skills, experience and/or training. Your recruiter can share more information about the specific salary range for your desired work location during the hiring process. We want our employees and their families to