LLM Solutions Architect

Design end-to-end agentic architectures — tool-use schemas, intent parsing, multi-step orchestration, and safety guardrails — engineered for long-term ownership by product engineering teams, not solo maintenance.

Define the multi-modal interface strategy across our product portfolio: how the same capability is exposed via UI, API, SDK, and agentic natural language — consistently and without duplication.

Design the horizontal LLM platform layer — shared RAG pipelines, prompt libraries, vector search infrastructure, and evaluation frameworks — that product engineering teams can build on and operate independently.

Prototype rapidly to validate AI product hypotheses before full engineering investment. Prototype acceptance by product teams is a primary success signal.

Ensure every system you architect comes with the observability, documentation, and engineering runbooks needed for a product squad to take ownership confidently.

Shape product strategy alongside Product leadership: actively influence what AI capabilities get prioritized, in what order, and with what trade-offs.

Select and govern LLM providers and deployment strategies per use case — balancing cost, latency, accuracy, and privacy requirements.

Drive alignment across Engineering, Product, and Design on what 'agent-ready' means for each product surface.

Mentor engineers on LLM integration patterns, agent evaluation, and production deployment practices — building the team's capability to own what you design.

5+ years of engineering experience, with at least 2 years designing and deploying LLM-powered systems in production.

Proven track record designing agentic systems: tool-use, function calling, multi-step reasoning, orchestration, and error recovery at production scale.

Experience designing AI systems for engineering team ownership — including observability standards, handoff documentation, and runbooks that let other teams maintain what you build.

Hands-on experience with major LLM APIs (OpenAI, Anthropic, Google Gemini) and at least one open-source model stack.

Experience building RAG pipelines with vector databases and orchestration frameworks (LangChain, LlamaIndex, or custom).

Strong Python engineering skills — production-grade LLM services, not just notebooks.

Demonstrated ability to influence product direction: you have shaped what gets built, not just how.

Clear communication in both directions: architectural trade-offs to engineers, business outcomes to executives.

Background in gaming, payments, or e-commerce — understanding of developer workflows, monetization models, or merchant operations.

Fine-tuning experience (PEFT/LoRA) for domain-specific model adaptation.

Experience with multi-agent orchestration frameworks (AutoGen, CrewAI, or custom).

Familiarity with LLM evaluation frameworks (RAGAS, DeepEval, or custom harnesses).

Exposure to EU AI Act, GDPR, or other AI compliance frameworks.