AI Engineer (Palantir AIP Specialist)

As an AI Engineer specializing in Palantir AIP, he will be responsible for the "intelligence layer" of the POC. While the data pipelines and Ontology provide the foundation, his role is to build the Generative AI workflows, Agentic logic, and LLM-powered applications that allow users to interact with Syteline ERP data using natural language and automated reasoning.

Core Responsibilities

1. AIP Logic & Tool Building

• AIP Logic Development: Use the AIP Logic suite to build functions that allow LLMs to query the Palantir Ontology (e.g., "Find all delayed work orders for Part X").

• Prompt Engineering: Design, test, and optimize complex prompts within the AIP Prompt Console to ensure high-accuracy responses from LLMs (GPT-4, Claude, etc.) using enterprise data.

• Tool Configuration: Build "Tools" that the AI can use to perform actions, such as summarizing supply chain risks or suggesting reorder quantities based on historical Syteline data.

2. Agentic Workflow Orchestration

• AIP Agents: Develop autonomous or semi-autonomous AI Agents that can monitor Syteline data changes (e.g., a sudden inventory drop) and proactively alert stakeholders with recommended solutions.

• RAG Architecture: Implement Retrieval-Augmented Generation within Foundry to ensure the AI's answers are grounded strictly in operational data, minimizing hallucinations.

3. Application & UX Delivery

• Conversational Interfaces: Build LLM-powered chat interfaces within Palantir Workshop that allow operations managers to "talk to their ERP."

• Decision Support: Design "Sidecars" and AI-driven widgets that provide real-time suggestions to users as they navigate the Supply Chain Command Center.

4. AI Governance & Accuracy

• Output Validation: Establish "Golden Datasets" to evaluate the accuracy of AI suggestions against known Syteline business rules.

• Guardrail Implementation: Ensure the AI adheres to security permissions defined in the Ontology (e.g., an AI agent should not reveal sensitive pricing data to unauthorized users).

Key Skills & Qualifications

Technical Must-Haves:

• Palantir AIP Proficiency: Hands-on experience with AIP Logic, AIP Prompt Console, and AIP Features in Workshop.

• LLM Orchestration: Deep understanding of how LLMs interact with structured data (Ontology) and unstructured data (PDFs/Manuals).

• Python for Data Science: Strong ability to write Python functions for custom logic, data post-processing, and API integrations.

• Semantic Modeling: While you won't build the Ontology, you must deeply understand how to navigate an Object-Oriented data model to "point" the AI at the right information.

Preferred Skills:

• Experience with LangChain, Semantic Kernel, or similar AI orchestration frameworks.

• Basic understanding of Supply Chain/Operations logic (Lead times, Safety stock, BOMs).