Job Title: AI Knowledge & Execution Grounding Specialist<\/span><\/b>
<\/p>This person is the architect of the\n"Ground Truth"<\/b> that ensures the AI\u2019s advice is high\-velocity\nand hyper\-practical.<\/span>
<\/p>Role Summary<\/span><\/b>
We are looking for a Knowledge & Execution Grounding Specialist<\/b> to\nensure our AI provides high\-growth SMBs with accurate,\n"ready\-to\-execute" strategies. You will sit in the Knowledge Team and\nact as the bridge between our proprietary growth frameworks and the\nTechnical/Product teams. Your mission is to structure our "execution\nblueprints"\u2014templates, SOPs, financial models, and growth hacks\u2014so the AI\nretrieves and applies them with 100% factual accuracy. You aren't just managing\ncontent; you are managing the logic of execution<\/b>.<\/span><\/p>Key Responsibilities<\/span><\/b>
<\/p>- Content Engineering for RAG<\/span><\/b>: Audit\n and restructure unstructured data (PDFs, wikis, transcripts) into\n "AI\-ready" formats by optimizing semantic chunking<\/b> and\n hierarchical data decoding.<\/span>
<\/li>- Blueprint Deconstruction:<\/span><\/b> Break\n down complex growth strategies (e.g., "Scaling Sales Teams" or\n "Inventory Management") into modular, machine\-readable\n blocks<\/b> that the AI can accurately recompose for a user\u2019s specific\n business size and industry.<\/span>
<\/li>- Operational Metadata Design:<\/span><\/b> Develop\n a tagging system that accounts for execution constraints<\/b>.\n (e.g., Tagging advice by Capital Required<\/i>, Team Size<\/i>,\n or Tech Stack<\/i> so the AI doesn't suggest a solution the\n SMB cannot execute).<\/span>
<\/li>- RAG Strategy Liaison:<\/span><\/b> Partner\n with the Tech team to define retrieval logic<\/b>. You decide which\n "Playbooks" are the primary sources for specific user intents to\n prevent the AI from giving "generic" internet advice.<\/span>
<\/li>- Execution Auditing:<\/span><\/b> Conduct\n "Stress Tests" on AI outputs. If a business owner asks "How\n do I set up a CRM?", you ensure the AI pulls from our<\/i> verified\n partner stack and doesn't hallucinate a random software.<\/span>
<\/li>- Grounding Quality Assurance<\/span><\/b>:\n Regularly audit AI outputs for hallucinations<\/b> and\n "unfounded" claims. Identify if errors stem from poor source\n data or technical retrieval failures.<\/span>
<\/li>- Contextual Guardrails:<\/span><\/b> Help refine\n the "Safety Logic" for the AI. If an SMB is in a pre\-revenue\n stage, you ensure the grounding layer prevents the AI from suggesting\n high\-burn growth strategies.<\/span>
<\/li><\/ul>Required Experience & Skills<\/span><\/b>
<\/p>- 3\u20135 Years Experience:<\/span><\/b> Ideally\n in Management Consulting<\/b>, Operations<\/b>, or Business\n Analysis<\/b> within the SMB/Startup ecosystem.<\/span>
<\/li>- AI Grounding Literacy:<\/span><\/b> You\n must understand how Retrieval\-Augmented Generation (RAG)<\/b> works\u2014specifically\n how "chunking" and "vector embeddings" impact the\n quality of a business recommendation.<\/span>
<\/li>- Process Mapping:<\/span><\/b> Proficiency\n in tools like Miro, LucidChart, or Notion to map out "Execution\n Workflows" before they are fed into the AI.<\/span>
<\/li>- Technical Familiarity & Specific AI Tooling: <\/span><\/b>While this is a Knowledge Team role, you will use JSON,\n YAML, and SQL<\/b> as the 'instructional languages' to define how our\n AI orchestration tools (LlamaIndex or LangChain<\/b>) interact with our\n proprietary growth content. You will not write application code, but you\n will be responsible for structuring our execution blueprints into these\n formats so the AI can filter, prioritize, and retrieve the exact right\n 'action step' for an SMB\u2019s specific context (e.g., industry, team size, or\n revenue stage) without hallucinating.<\/span>
<\/li>- The "Scrappy" Mindset:<\/span><\/b> A\n builder who can manually audit a 50\-step growth plan and identify exactly\n where the AI lost the "logic" of the execution.<\/span>
<\/li><\/ul>
<\/span>
<\/div> <\/span>
<\/p> <\/span>
<\/p>
<\/div><\/span>
