AI Solutions Engineer

Location: San Francisco, CA, US; Remote, US

Department: Engineering, Product and Design (L2)

We're building a new capability at Pinterest: embedding AI-native engineering directly inside our business functions. The AI Solutions Engineer will partner with teams across Marketing, Finance, Sales, HR, Legal, and other functions to surface high-value automation opportunities, then design and ship the AI-powered tools that bring those opportunities to life.

This is a hands-on, mid-level software engineering role for someone who is equally comfortable reading a business process flowchart and writing production-grade Python. You'll work end-to-end — from discovery and scoping through prototyping, launch, and iteration — using the latest agentic frameworks, tool-calling patterns, and responsible AI practices.

What you'll do:

• Discover and scope AI opportunities: Partner with internal teams across corporate functions to understand their workflows, pain points, and goals, and identify high‑value AI/automation opportunities. Map and improve business processes: document current workflows, identify bottlenecks, and propose AI‑enabled changes that deliver clear business outcomes (e.g., time or cost savings, improved quality or compliance).
• Design end-to-end AI solutions: Design and implement AI‑enabled tools and workflows that integrate with existing systems and data sources, and that are intuitive for non‑technical users.
• Build and ship production-quality software: Write clean, maintainable code and tests. Use standard CI/CD and environment practices. Implement logging, monitoring, and basic guardrails so we can understand and improve performance, quality, cost, and reliability over time.
• Pilot, rollout, and drive adoption: Pilot, roll out, and drive adoption of solutions by working closely with end‑users, gathering feedback, and iterating based on real‑world usage.
• Champion for responsible AI: Ensure solutions follow privacy, security, and compliance expectations, especially when working with sensitive or regulated data.
• Build for reuse: Create and share reusable patterns, components, and documentation to make future AI/automation work faster and more consistent across teams.
• Accelerate Workflows with Generative AI and Automation: Leverage AI to accelerate execution (e.g., draft, prototype, outline), explore alternative solutions using AI (iterate, compare approaches), synthesize information with AI (summarize, distill key themes), automate repeatable work (documentation, reporting, QA checks)

What we're looking for:

We're looking for mid-level engineers who have already shipped something real with AI — and who can work as a peer with non-technical business partners, not just as an order-taker. Specifically, you bring:

• Software engineering foundation. A CS, Engineering, Data Science, or related degree (or equivalent experience), with demonstrated ability to build and operate production systems — backend services, internal tools, integrations, or data applications.
• Hands-on AI and automation delivery. You've shipped AI-powered or automation-driven solutions in a real environment. Examples include: a multi-step workflow automation, an internal tool using document understanding or intelligent routing, or an integration of an AI service (e.g., OpenAI, Anthropic, Vertex AI, Bedrock) into an existing system.
• Agentic AI literacy. You understand how modern agentic systems are constructed — the difference between local and remote agents, how MCP (Model Context Protocol) works, what Agent Skills and Hooks are for, and how A2A (Agent-to-Agent) coordination is structured. You can reason about when to use these patterns and when simpler approaches suffice.
• System design and architecture thinking. You can sketch a data flow, reason about integration points, evaluate trade-offs between approaches, and design for failure — including fallbacks, retry logic, timeouts, and human escalation paths.
• Data and security judgment. You understand data access controls, the risks of giving AI broad access to sensitive information, PII minimization, audit logging, and what responsible data handling looks like in an enterprise environment. You know to filter data before it reaches the model, not after.
• Business function acumen. You can engage credibly with stakeholders in Marketing, Finance, Sales, HR, Legal, or Operations — understanding their workflows, KPIs, and constraints well enough to scope solutions that fit their real needs.
• Clear, collaborative communication. You can explain architecture trade-offs to a Finance Manager and debug a prompt failure with an engineer in the same afternoon. You build trust by being direct about limitations, not by over-promising.
• Preferred Qualifications
  • Experience working embedded with or alongside corporate / G&A functions (Finance, Legal, HR, Marketing, Sales Operations, or similar).
  • Practical experience with agentic frameworks such as LangGraph, Claude Agent SDK, or comparable tooling
  • Familiarity with MCP server design — including building, deploying, and securing MCP-compliant tool servers.
  • Experience designing and evaluating AI outputs at scale: eval sets, sampling pipelines, human-in-the-loop review queues, or A/B testing of AI-powered features.
  • Exposure to responsible AI frameworks: data minimization, differential privacy concepts, model output auditing, or working in PII-sensitive / regulated domains.
  • Experience with RAG (Retrieval-Augmented Generation) pipelines, vector databases, or enterprise search integrations.
  • Familiarity with CI/CD for AI: prompt versioning, model version pinning, regression testing for LLM-powered features.
  • Bachelor’s/Master’s degree in a relevant field such as Computer Science or equivalent experience.
  • Demonstrated ability to use AI to improve speed and quality in your day-to-day workflow for relevant outputs.
  • Strong track record of critical evaluation and verification of AI-assisted work (e.g., testing, source-checking, data validation, peer review).
  • High integrity and ownership: you protect sensitive data, avoid over-reliance on AI, and remain accountable for final decisions and deliverables.

Relocation Statement:

• This position is not eligible for relocation assistance. Visit our PinFlex page to learn more about our working model.
  
  

In-Office Requirement Statement:

•  This role will need to be in the office for in-person collaboration 1-2 times every 6-months and therefore can be situated anywhere in the country.
  
  

#LI-REMOTE

#LI-KBF