--- title: 'Member of Technical Staff — Pretraining Infra (Experienced) at Nuance Labs' canonical: 'https://feeny.ai/job/member-of-technical-staff-pretraining-infra-experienced-nuance-labs-seattle-vbdypnk5fzcv' type: 'job' last_seen: '2026-09-11' --- # Member of Technical Staff — Pretraining Infra (Experienced) at Nuance Labs - **Company:** Nuance Labs - **Location:** Seattle, WA - **Posted:** 2026-06-05 - **Last confirmed live:** 2026-09-11 - **Apply:** https://job-boards.greenhouse.io/nuancelabs/jobs/4274385009 ## Job description ## About Nuance Labs Nuance Labs is building photorealistic, real-time AI avatars with emotional intelligence: a full-duplex audiovisual system that can listen, speak, react, interrupt, and respond like a real person. We're a research company, with PhDs from MIT, UW, Oxford, CMU, and Johns Hopkins, and industry experience from Apple, Meta, Amazon AGI, and more. Backed by Accel, Lightspeed, South Park Commons, and NVIDIA, we combine frontier research with ruthless engineering needed for consumer-grade, real-time systems. The team is small, the work is real, and the problems are unsolved. How Nuance Differentiates Most conversational AI avatars today are hacks — a face slapped on a speech-to-speech pipeline, stuck in the uncanny valley: emotionless, mechanical, one-turn-at-a-time. Current systems take 2–5 seconds to respond; natural conversation requires sub-500ms. That's a 10x improvement, and it demands rethinking the entire stack. That rethinking starts with full-duplex: an AI that listens and speaks simultaneously, perceives emotion in real time, and responds with a face that actually reflects it. It's an extremely hard problem, and we're developing foundation models designed for it from the ground up. ## About the Role We're looking for a deeply technical MTS to own distributed training infrastructure for large-scale omni model pretraining. This role sits at the intersection of research, systems, and GPU-scale execution — building the training stack from 0→1 and scaling it: distributed execution, parallelism, GPU communication, data loading, checkpointing, observability, and debugging. Our models are omni from the ground up (audio, video, language, real-time full-duplex), which introduces systems challenges beyond standard LLM training: multimodal synchronization, long temporal context, variable sequence lengths, and tight memory/throughput constraints. High ownership. Direct impact on what models we can train, how fast research can iterate, and how reliably we scale. ## What You’ll Own - Own the distributed training stack for omni model pretraining, from 0→1 system design to 1→10 scaling across large GPU clusters. - Build and operate the core training runtime: job orchestration, distributed execution, checkpointing, recovery, monitoring, and debugging for long-running training jobs. - Optimize large-scale training performance across parallelism strategy, GPU communication, memory usage, data throughput, MFU, step time, and end-to-end training efficiency. - Build infrastructure for omni training workloads: high-throughput audio/video/text data loading, temporal alignment, variable sequence handling, multimodal synchronization, and memory-efficient training. - Evolve the platform as model architectures, training recipes, data mixtures, sequence lengths, hardware constraints, and research directions change. ## What We’re Looking For - Hands-on experience running large-scale distributed training jobs across large GPU clusters; experience at hundreds of GPUs minimum, 1,000+ GPUs a strong plus. - Deep understanding of distributed training mechanics: data/tensor/pipeline/sequence parallelism, gradient communication, collectives, mixed precision, activation checkpointing, optimizer state, memory pressure, and framework-level tradeoffs. - Strong understanding of GPU communication and performance debugging: NCCL, all-reduce/all-gather/reduce-scatter, communication-computation overlap, topology, synchronization, stragglers, low MFU, OOMs, checkpoint bottlenecks, and data starvation. - Practical experience with at least one major large-scale training stack such as Megatron, PyTorch FSDP, DeepSpeed, or equivalent internal infrastructure. - Understanding of omni or multimodal training challenges, especially audio/video/language data, long temporal context, variable sequence lengths, modality-specific bottlenecks, and high-throughput dataloading. - Strong software engineering fundamentals, curiosity, and adaptability to new model architectures, training frameworks, hardware constraints, and research ideas. Bonus Points - Prior 0→1 experience building large-scale training infrastructure or deeply modifying core training frameworks, runtimes, checkpointing, or debugging systems. - Experience training large omni or multimodal models involving audio, video, text, or long-context temporal data. - Experience with adjacent infrastructure areas such as RL/post-training, data infrastructure, synthetic data generation, evaluation, or serving. - Publications or substantial open-source contributions in ML systems, distributed systems, HPC, GPU performance, or training infrastructure. ## Compensation $300,000 – $400,000 base salary, plus meaningful equity. We think long-term ownership matters and structure equity accordingly. Logistics - Location: In-person in Seattle, five days a week — we believe in the compounding value of working shoulder-to-shoulder. - Visa sponsorship: We sponsor visas (O-1, H-1B, green card, etc.) from day one. - AI-native tooling: Do your best work with the best tools, including unlimited tokens. ## Benefits - Health: HSA plan with ~$2,000 in annual company contributions — roughly 2x what most big tech companies put in. - Time off: 15 days of PTO plus public holidays, and we close the office for a full week at year-end. - Food: Lunch, drinks, and snacks on us every workday — the small thing that quietly makes the day better. - Commuter benefits: We help cover the cost of getting to the office. - 401(k) Nuance Labs is an equal opportunity employer. We believe diverse teams build better AI. ## About Nuance Labs ## Company Overview - **One-liner**: Nuance Labs is building a real-time, full-duplex audiovisual AI model that can see, hear, reason, speak, and express emotion simultaneously, enabling face-to-face interaction with machines. - **Entity Type**: Private (Seed stage; raised $10M in a seed round led by Accel) - **Headquarters**: Seattle, Washington, United States - **Founded**: Not publicly disclosed (likely 2024 or 2025 based on funding date) - **Founders**: Fangchang Ma (CEO), Edward Zhang (CTO), Karren Yang (Chief Scientist) – all PhDs from MIT with prior experience at Apple ## Core Business - Primary industry: Artificial Intelligence / Software Development - Target customers: B2B (enterprise and consumer product integration) and potentially B2C (direct-to-consumer AI interfaces) - Mission or purpose statement: “We are creating a world where people can finally talk to every product, face to face.” – building a “human foundation model” with social and emotional intelligence. ## Products & Services - **The Full-Duplex Engine**: A real-time audiovisual model that perceives, reasons, and responds within 500ms – handling interjections, nodding, backchanneling, tone, and expression. The system is designed to cross the uncanny valley with warm, engaging personality. - **Human Foundation Model**: An AI that learns human behavior by predicting the next audio and visual token (similar to how transformers predict the next word). It reads tone, expression, hesitation, and responds in real time. ## Market Standing - **Valuation**: Not disclosed (private company) - **Key Metric**: Total funding – $10M in a seed round closed on October 9, 2025 - **Notable Investors/Partners**: Lead investor Accel (with 3 total investors in the seed round) - **Growth Signals**: - Small team of ~10 employees with 9.1% monthly headcount growth - Strong hiring momentum: 6+ active job postings for research and engineering roles - Team includes PhDs from MIT, UW, Oxford, CMU, Johns Hopkins, and alumni from Apple, Meta, Amazon AGI, and Discord - High LinkedIn follower growth (+4.5% monthly) ## Competitive Advantages - **Full-duplex real-time capability**: 500ms latency floor – a hard requirement for natural conversation that most current AI systems cannot achieve. - **Multimodal emotional intelligence**: Reads tone, expression, and hesitation, not just text. - **Elite founding team**: PhDs from top institutions with deep experience shipping ultra-low-latency ML products at Apple and other tech giants. - **First-mover in “human foundation model”**: Applying next-token prediction to audio-visual human behavior, not just language. ## Strategic Focus - Current priorities: Advance the Full-Duplex Engine, scale the model, and build the infrastructure for real-time audiovisual interaction. The company is hiring across pretraining infrastructure, RL research, model optimization, speech synthesis, video diffusion, and MLLM training. - Long-term direction: Enable face-to-face AI interaction for every product, moving computing interfaces closer to human communication. ## Why Work Here - **Culture**: “Small, fast-moving research team” with an “exceedingly high bar” – only the very best talent. Every member has massive ownership, deep trust, and the opportunity to shape both the technology and the company from the ground up. - **Location**: Based in Seattle, WA. The company appears to be primarily in-office (all listed roles are in Seattle, and the team is small). - **Notable perks**: Not explicitly mentioned, but the emphasis on research autonomy, high-impact work, and a founding team from top AI labs suggests a strong engineering culture with cutting-edge projects. - **Open roles** (as of mid-2026): Research Scientist (Speech Synthesis, Video Diffusion, MLLM Training), Machine Learning Research Engineer, ML Infra Engineer, Systems Engineer (Real-Time Engine), AI Product Engineer, and more. ## Sources 1. [nuancelabs.ai](https://www.nuancelabs.ai/) 2. [nuancelabs.ai/careers](https://www.nuancelabs.ai/careers) 3. [nuancelabs.ai/about](https://www.nuancelabs.ai/about) 4. [job-boards.greenhouse.io/nuancelabs](https://job-boards.greenhouse.io/nuancelabs) 5. [linkedin.com/company/nuance-labs-ai](https://www.linkedin.com/company/nuance-labs-ai) ## Other roles at Nuance Labs - [Human Evaluation Researcher](https://feeny.ai/job/human-evaluation-researcher-nuance-labs-seattle-gvf6yh22ba2e) — Seattle, WA - [Nuance Research Fellow](https://feeny.ai/job/nuance-research-fellow-nuance-labs-seattle-hge7hy4qpd3t) — Seattle, WA - [Member of Technical Staff — RL Research (New PhD Grad)](https://feeny.ai/job/member-of-technical-staff-rl-research-new-phd-grad-nuance-labs-seattle-wmvfxyd7gw98) — Seattle, WA - [Member of Technical Staff — Model Optimization and Inference (New Grad)](https://feeny.ai/job/member-of-technical-staff-model-optimization-and-inference-new-grad-nuance-labs-ech27rtf16gn) — Seattle, WA - [Member of Technical Staff — ML Infra (Data)](https://feeny.ai/job/member-of-technical-staff-ml-infra-data-nuance-labs-seattle-efnr92fwhx14) — Seattle, WA - [Member of Technical Staff — Model Optimization and Inference (Experienced)](https://feeny.ai/job/member-of-technical-staff-model-optimization-and-inference-experienced-nuance-5gazzjwgrhm1) — Seattle, WA - [Member of Technical Staff — RL Research (Experienced)](https://feeny.ai/job/member-of-technical-staff-rl-research-experienced-nuance-labs-seattle-r895bhr1agh6) — Seattle, WA