AI Solutions Architect

Role overview Translate messy business constraints into clear AI system designs that a delivery team can actually build. The role favors pragmatic, minimal architectures over ambitious ones, weighing cost, latency, accuracy, and risk to land on the simplest shape that can do the job reliably.

Responsibilities - Lead architecture conversations with stakeholders: scope the problem, frame the tradeoffs, and recommend an approach. - Produce system diagrams, reference implementations, and written design documents that engineers can execute against. - Evaluate models, vendors, and integration patterns, including build-versus-buy decisions for retrieval, orchestration, and evaluation tooling. - Define delivery plans with milestones, dependencies, and rough effort estimates. - Stay close enough to the code to know whether a design is realistic, without becoming the primary implementer.

Requirements - Significant experience designing and shipping AI or ML systems in production, not just prototypes. - Strong working knowledge of LLM patterns: prompting, fine-tuning, retrieval-augmented generation, tool use, evaluation, and guardrails. - Comfort with at least one major cloud platform and modern data or ML infrastructure, including vector stores, orchestration, and observability. - Clear written and verbal communication; able to explain a tradeoff to an executive and an engineer with equal clarity. - Pragmatic decision-making: knows when to recommend a quick notebook demo and when to insist on a full platform.

Nice to have - Background in consulting, solutions engineering, or technical pre-sales. - Familiarity with regulated or high-stakes domains where architecture choices carry outsized risk.