# Senior Principal AI Solutions Engineer — SambaNova Systems

- Location: San Jose, CA
- Posted pay: $234k–$286k/yr
- Posted: Sep 29, 2026
- Status: open (checked Nov 21, 2026)
- Apply (employer's site): https://sambanova.ai/sambanova-available-positions/?gh_jid=6211407004
- Employer record: https://visajobs.io/company/sambanova-systems.md

## Sponsorship history
- Tier: Employer history only
- With this exact title: 0

## Description
SambaNova is a leader in next-generation AI infrastructure, delivering a full-stack inference platform for customers worldwide. At the core of SambaNova's technology is the RDU (Reconfigurable Dataflow Unit) — a chip built on a dataflow architecture rather than the traditional GPU model. Its decode performance is especially strong for agentic workloads like multi-turn agents, code generation, and long-running applications. RDUs are packaged into SambaRack, rack-scale hardware that lets customers deploy state-of-the-art models with better performance, greater energy efficiency, and faster time to value.
We are looking for a Senior Principal AI Solutions Engineer: a hands-on technical leader who sets the technical direction for our solutions portfolio and still builds. This is someone who can go from a customer problem to a polished, working, agentic application in days, and then harden it for production. You will work across the whole modern AI stack: open-weight models, agent frameworks and tool use, self-improving evaluation loops, multimodal and voice pipelines, and the front end that makes it all usable by knowledge workers. You will lead hands-on engagements with our most strategic customers, mentor and raise the bar for the wider solutions team, and feed what you learn back into the platform.
Responsibilities
Build agentic and self-improving systems
Agentic AI Architecture: Design and build single- and multi-agent systems with planning, tool and function calling, MCP, memory, and long-running workflows that are reliable enough for production • Self-Improving Systems: Build feedback loops that make agents better over time, including automated evals, LLM-as-judge, trace-driven prompt and program optimization (e.g. DSPy-style), synthetic data generation, and fine-tuning from production signals
Evaluation and Observability: Set up eval harnesses, tracing and guardrails so customers can measure quality, cost and latency, and trust what they deploy
Get the most from open models
Open-Weight Model Expertise: Choose, adapt and combine open models (e.g. Llama, Qwen, DeepSeek, gpt-oss, Gemma, GLM) for customer use cases, including LoRA/PEFT fine-tuning, distillation and model routing
Performance and Benchmarking: Benchmark end-to-end solutions, not just tokens per second, and show where fast inference changes what an application can do
Model Integration: Validate new models and capabilities on SambaNova's platform stack and report gaps to Product and Engineering
Ship full-stack vertical solutions
Front-End and Product Build: Build polished, usable web applications (e.g. React/Next.js, TypeScript) that non-technical knowledge workers can adopt, not just notebooks and APIs
Vertical Knowledge Work: Build domain solutions for areas such as financial services, legal, healthcare, public sector and research: document analysis, research agents, RAG and knowledge assistants, report generation and workflow automation
Multimodal Solutions: Build vision-language and document-understanding pipelines (OCR, charts, forms, images, video) combined with agentic reasoning
Audio and Voice: Build real-time voice agents and audio pipelines (ASR, TTS, speech-to-speech, streaming and turn-taking) where low latency is the product
Lead with customers and shape the platform
Solutions Technical Strategy: Set the technical direction for the solutions portfolio: which agentic, multimodal and vertical patterns we invest in, the shared frameworks and components we build, and the engineering standards they meet
Technical Leadership and Mentorship: Act as the senior technical authority across solutions engineering; review designs, mentor engineers and lift the quality of everything the team ships
Executive Engagement: Act as the trusted technical advisor to customer CTOs and AI leaders on architecture, build-versus-buy and scaling AI across the enterprise
Customer Engagement: Lead technical discovery, workshops, hackathons and hands-on co-builds with strategic customers, from prototype to production
Reference Architectures: Publish reference architectures, starter kits, open-source examples and best-practice guides that scale beyond each engagement
SambaStack Tooling: Develop tooling and automation for SambaStack deployment, management and integration
Product Feedback: Feed what you learn in the field (model requests, feature gaps, developer experience) into the model roadmap and platform priorities
Thought Leadership: Represent SambaNova through demos, talks, blogs and community contributions
Requirements
Required Qualifications
Bachelor's degree or higher in Computer Science, Electrical Engineering, Applied Mathematics, Physics, Statistics or a related field
8+ years (IC5) or 10+ years (IC6) of industry experience in software, ML or solutions engineering, including 3+ years building LLM-based applications that reached production
Proven experience building agentic systems: tool and function calling, multi-agent orchestration and frameworks such as LangGraph, CrewAI, OpenAI Agents SDK, Claude Agent SDK or equivalent • Hands-on experience with open-weight models: serving, prompting, fine-tuning (LoRA/PEFT) and evaluation • Strong full-stack skills: expert Python plus modern front-end development (TypeScript, React/Next.js), APIs and cloud deployment
Experience designing evals and benchmarks for LLM applications, covering quality, latency and cost • Excellent customer communication, with the ability to lead workshops and explain architecture to both executives and engineers
A track record of technical leadership across teams: setting architecture direction, mentoring senior engineers and owning outcomes for strategic accounts
Preferred Qualifications
Experience building self-improving or learning systems: prompt/program optimization, RL from feedback, synthetic data pipelines or continual fine-tuning
Experience building real-time voice agents (e.g. LiveKit, Pipecat, WebRTC) and working with speech models (ASR/TTS)
Experience wit

> History, not a promise: past H-1B filings do not mean an employer will sponsor you or a given role. Data: U.S. DOL OFLC LCA disclosure (FY2024–FY2026 for role counts, certified LCAs) and the USCIS H-1B Employer Data Hub. Not legal advice. Methodology: https://visajobs.io/methodology
