Senior AI Engineer (Edge Dialog Systems)
BrightAI Corporation3 hours ago
Palo Alto, CA, USASenior
Responsibilities
- Own the on-device dialog pipeline, including intent routing, hybrid intent classification, speech-input normalization, and multi-step guided procedures.
- Maintain deterministic safety guardrails around language-model output, including confirmation gating, criticality tagging, echo-back validation, and negation handling.
- Run and optimize small language models on constrained devices within memory, latency, computational, and thermal budgets.
- Maintain the configuration-driven model for adding device commands and customer procedures without code releases.
- Coordinate device deployment with the edge team and maintain the API contract with the on-device voice and speech-to-text stack.
- Define and run benchmarks for latency, accuracy, and safety-critical false-accept and false-reject rates.
- Build golden datasets and non-regression suites used as release gates as command and procedure catalogs expand.
- Lead migration from zero-shot to fine-tuned on-device models while avoiding per-customer retraining.
- Collaborate with product, firmware, cloud, and domain teams to launch languages, device commands, and guided workflows.
Requirements
- 5+ years of experience in ML or AI focused on NLP, LLMs, or conversational AI.
- Applied experience with NLU, dialog systems, or on-device machine learning.
- Strong experience with LLM prompting, structured output, tool and function calling, evaluation, embeddings, semantic similarity, and retrieval-augmented generation.
- Strong Python skills and fluency with pytest, Git, and tested, reviewable production code.
- Experience building edge conversational systems with multi-turn dialog management and efficient intent/NLU pipelines.
- Experience implementing disambiguation, repair, deterministic guardrails, confirmation gating, negation handling, and state-machine or workflow engines.
- Experience deploying models to constrained hardware such as NPUs, mobile devices, or embedded targets, including ONNX, ONNX Runtime, quantization, and aarch64 packaging.
- Practical embedded development experience with Linux, Docker, adb, systemd services, and device-log diagnostics.
- Experience taking ownership of and maintaining a complex existing codebase.
- Strong benchmarking, dataset, regression-testing, problem-solving, written-communication, and cross-functional collaboration skills.
- Preferred experience with SLM fine-tuning, LoRA, QLoRA, instruction and format tuning, distillation, INT8 and INT4 quantization, ONNX optimization, profiling, and data-flywheel development.
- Bonus qualifications include speech recognition, LLM-as-a-judge evaluation, multilingual NLU, industrial or safety-critical products, Go, MCP, agentic tooling, and startup experience.
Benefits
- Full-time position based in Palo Alto, California.
- On-site or hybrid work arrangement.
Categories
About BrightAI Corporation
BrightAI is transforming essential services with Physical AI—real-world intelligence that drives proactive, data-driven operations. Built on the Stateful platform, BrightAI empowers operators of critical infrastructure to collect and connect sensor data in real time, uncover hidden insights, and make smarter decisions. From predictive diagnostics and autonomous robotics to digital twins and AI-enabled workflows, BrightAI solutions serve industries including water, power, gas compression, pest control, HVAC, and manufacturing. By turning complex physical signals into actionable intelligence, BrightAI is redefining how essential services are delivered and sustained.