Applied AI Engineering Lead

Thresh Consulting is hiring an Applied AI Engineering Lead to set technical direction and build production AI solutions for clients, with a focus on applied LLM engineering and customer-facing experiences. The role leads a small engineering team, partners with client and practice leaders, and owns AI engagements from discovery through production deployment.

Responsibilities:

• Set technical direction for AI engagements
• Partner with clients to identify and prioritize high-value AI use cases, and be clear about what's ready for production versus what's still a demo
• Architect end-to-end LLM solutions: model selection, prompting strategy, retrieval, agents and tool use, and the integrations that connect them to real systems
• Make build, buy, prompt, or fine-tune calls based on cost, latency, quality, and risk, and explain those calls to client leadership
• Shape scope, estimates, and technical approach for new AI engagements alongside Thresh's practice and sales leads
• Build conversational shopping experiences where customers discover, compare, and buy through natural language, from first question to completed order
• Set the patterns for streaming interfaces that stay fast and readable under real load, including partial responses, tool-call states, errors, and graceful fallbacks
• Bring real business logic into the conversation: live pricing, inventory, promotions, personalization, and customer account context
• Build for a client's own site and app and for third-party AI assistants like ChatGPT and Gemini, all from one shared foundation
• Build and extend MCP servers and tools that give models safe, structured access to catalog, pricing, cart, and order capabilities
• Prepare commerce systems for AI agents that shop on a customer's behalf, with structured product data, agent-ready checkout, and clear rules for what an agent can and can't do
• Work with emerging agentic commerce protocols (ACP, UCP), and know where each fits a client's business model
• Integrate with enterprise commerce, catalog, order, and identity systems that were never designed for AI
• Design multi-step agents and workflows that take real actions in client systems, with clear permissions, human-in-the-loop checkpoints, and audit trails
• Build retrieval and knowledge systems (RAG) over client content, catalogs, and documentation, and tune them for accuracy
• Build guardrails that keep the assistant in scope. That means no fabricated prices, promises, or commitments, and a clean handoff to a person or standard flow when a request goes past what it should handle
• Own the evaluation strategy. Build eval sets and automated test harnesses covering clean queries, edge cases, out-of-scope requests, and prompt-injection attempts, and turn what you find into fixes and release gates
• Instrument conversations end to end for analytics, attribution, observability, cost, and latency, so clients can measure what the experience drives
• Prepare work for the legal, privacy, security, and brand reviews that every transactional AI experience goes through
• Monitor model quality, drift, and spend after launch, and plan for model upgrades and vendor changes without breaking the experience
• Make AI coding agents (Claude Code, Cursor) your default way of working
• Write specs precise enough for an agent to implement against, then verify the output against the spec rather than the diff
• Prototype conversational features in days with mock data, so clients can see the experience before the real integrations land
• Test new tools and workflows, and bring what works back to the team and to the wider Thresh practice
• Mentor engineers through code review, pairing, and technical guidance, across the full applied AI stack, from prompting and evals to production architecture
• Partner with product managers and designers on conversation design, scope, technical risk, and estimates
• Define the components, patterns, and standards that become reusable assets across Thresh's AI work, including reference architectures and starter kits
• Help build Thresh's applied AI capability through hiring input, technical interviews, and internal enablement
• Represent Thresh's AI point of view with clients and platform partners through workshops, technical content, and pre-sales conversations
• Own AI solutions from discovery through production deployment
• Explain progress and tradeoffs clearly to technical and non-technical stakeholders, including what AI can and can't reliably do
• Work within each client's existing stack, frameworks, and constraints
• Hold the quality bar when deadlines tighten and requirements shift