Solution Architect (AI/LLM Inference)

THE ROLE

As a Solution Architect (AI/LLM Inference) at Baseten you will partner closely with Sales and customers to translate business needs into technical solutions, run technical discovery, and guide repeatable deployments and proofs of value for customers. This role is a great fit for entrepreneurial, customer-facing technical professionals who want a front-row view into how modern companies adopt AI at scale, and who enjoy working across technical discovery, solution design, demos, deployment scoping, and hands-on customer implementations, in close partnership with Sales and Engineering.

RESPONSIBILITIES

• Partner with Sales on customer discovery calls (most often second calls, occasionally first calls for large accounts).
• Lead demos and technical scoping to align on success criteria, architecture, and deployment approach.
• Own benchmarking and repeatable deployments, including:
• Handling standard deployment patterns and configurations across many modalities – LLMs, embeddings, image and video generation, VoiceAI, etc.
• Advising on tradeoffs like H100s vs B200s and latency-optimized vs throughput-optimized setups.
• Driving consistent “playbook” style deployments for common models and use cases.
• Become a power user of different runtimes such as vllm, sglang, and TRT-LMM and all the common configurations and tradeoffs between them
• Drive POC and project execution, including:
• Scoping POCs and keeping stakeholders aligned on timeline, deliverables, and next steps.
• Acting as the “ringleader” or project manager for POCs.
• Pulling in Forward Deployed Engineering (FDE) support when deeper or more complex technical work is needed.

REQUIREMENTS

• AI/ML background and the ability to credibly discuss AI/ML topics with technical stakeholders.
• Strong customer-facing communication skills, including the ability to run structured discovery and clarify ambiguous requirements.
• Technical depth to scope solutions, without needing to write production code.
• Ability to script and prototype as needed, including comfort “vibe coding” to move quickly in technical workflows.

NICE TO HAVE

• Experience running or supporting benchmarks for ML inference deployments.
• Familiarity with infrastructure tradeoffs relevant to inference performance and cost (for example GPU selection and latency versus throughput tuning).
• Experience serving as a cross-functional technical lead for customer POCs, including coordination across Sales and Engineering.