Machine Learning Engineer, Speech LLMs   You’d be an early member of a new speech LLM lab inside a profitable, bootstrapped company, with the job of training models that understand and generate speech natively.   About the company A profitable Voice AI company based in San Francisco. It builds hardware and software that capture what professionals say and hear, then turn it into notes and usable information. The business has a large global user base and is now setting up a core SpeechLLM lab, which this person would join as a founding hire.   The role You’ll build and train large-scale audio and speech models from the ground up. That could mean unified SpeechLLMs, advanced ASR, expressive TTS or generative audio architectures. One day you’re designing a new sequence modelling architecture, the next you’re debugging a distributed training cluster. The work runs across the whole stack, from signal processing and raw acoustic representations to foundation model training and optimisation for edge devices. You’ll own ambiguous problems and take them into production.   What we’re looking for A proven track record of building and training large-scale audio or speech models. Deep expertise in PyTorch or JAX, including optimising large distributed training runs, managing GPU memory and fixing performance bottlenecks. Comfort moving between research and engineering, and a real interest in AI systems that handle speech natively.   Any of the following is a bonus: Text LLM pretraining, instruction tuning or RLHF Neural audio codecs for streamable, high-fidelity audio Diffusion, flow matching or autoregressive models for speech and voice generation RL techniques such as RLHF or GRPO to improve conversational cadence, steerability and alignment Inference optimisation with serving frameworks like vLLM, TensorRT-LLM or SGLang Large GPU clusters, distributed training with FSDP or DeepSpeed, and Kubernetes   Salary and benefits Up to $270K, plus a performance bonus and equity. Healthcare for you and your dependents, including dental and vision, with a generous employer subsidy. 401(k) with company matching. Unlimited PTO plus 13 paid holidays.   Logistics Full-time, permanent. Based in San Francisco, hybrid, with a minimum of three days a week in the office.

Agentic RL researcher, post-training (San Francisco or remote)   About the company A well-funded voice AI company in San Francisco, building speech models after a recent raise. The research team is growing, with several researcher roles open at once, and is kept deliberately flat. This role reports to the Head of AI.   The role You’ll own the RL side of post-training the next text-to-speech model. A big part of how that model gets good is how it’s post-trained, which means building complex, text-based agent loop systems that generate the training signal for RL. You’ll design and run PPO, GRPO, DPO or on-policy loops on model weights, with the environments and rewards that go with them. You’ll also build pipelines that write their own questions and tasks, so training scales without constant manual data curation. Ablations show what moved behaviour, and you’ll work with the speech researchers so what you build in text carries over to audio.   What we’re looking for Hands-on RL or preference learning that trained model weights and changed behaviour in a way you can measure. You’ve built or led an RL pipeline that scales, ideally one that generates its own questions and tasks (curating an SFT dataset on its own isn’t what this role needs). You’ve done it at a real company, with real constraints, on chatbot-style or conversational agents. Bonus: audio or speech post-training, large-scale distributed training, and published or open-source work in LLM post-training.   Salary and benefits Up to around $250k depending on experience, plus equity. Healthcare, dental and vision.   Logistics Full-time, permanent. San Francisco hybrid or fully remote. Start date ASAP.

Machine Learning Researcher, Audio (TTS) The text-to-speech model already works really well. The job is the last 10% of post-training: the pronunciation errors and small issues that only show up live.   About the company A well-funded voice AI company in San Francisco that builds AI phone agents for enterprise customers and now trains its own speech models. The research team has just finished evaluating its newest model and is moving into full development. Research headcount is growing, and text-to-speech is now the main hiring priority.   The role You’ll own the text-to-speech model from research prototype to production. That means post-training it until it’s reliable and stable, fixing mispronunciations and the issues that only appear live, and building on neural audio codecs and discrete audio representations.   You’ll ship into a low-latency, telephony-class path and work alongside the speech-to-speech and LLM researchers. You report to the Head of AI in a deliberately flat team, with a possible technical lead path for the right person.   What we’re looking for Hands-on ownership of a neural audio codec or audio tokenizer, from training objective through to evaluation. Experience training generative audio models with a clear personal contribution to the architecture and training, beyond data, evals or product glue. Real industry experience taking a TTS model from research prototype to production scale, plus post-training work that made it reliable and stable. You can ship as well as publish. Shipping into a low-latency telephony path, ASR experience alongside TTS, and first or second author papers in non-generic areas are a bonus.   Salary and benefits Up to around $300K depending on experience, plus equity. Healthcare, including dental and vision.   Logistics San Francisco office hybrid or fully remote. Start as soon as possible.

London | Remote You’ll take ownership of the product that helps businesses build, train and deploy machine learning models at scale. This is an opportunity to lead product direction within a growing AI infrastructure business. You’ll define the longer-term vision, decide what gets built and work closely with engineers, ML specialists and customers to improve how AI models are developed and operated. The product covers some of the biggest challenges in machine learning infrastructure: how models are trained and fine-tuned, how GPU resources are used, and how models are deployed and served efficiently in production. You’ll have genuine product ownership. Rather than simply managing a delivery backlog, you’ll shape the roadmap, identify opportunities to improve the platform and turn complex technical and customer requirements into clear product decisions. You’ll also work closely with engineering and customers, gaining direct insight into how organisations are building and deploying AI and using that feedback to influence the future direction of the platform. This could suit you if you’re already a Technical Product Manager working on ML platforms, cloud infrastructure, MLOps or developer tooling and want to take ownership of a product in a fast-moving AI market. You’ll need: 5+ years of Product Management experience, ideally owning technical platform or infrastructure products Experience with ML platforms, MLOps, AI infrastructure or cloud-based developer products A good understanding of machine learning workflows, including model training, deployment and inference Knowledge of technologies such as Kubernetes, PyTorch, Ray, Slurm or similar Experience defining product strategy, building longer-term roadmaps and working closely with engineering teams and technical customers A strong technical understanding through education or hands-on experience Experience with GPU infrastructure, distributed computing or model serving would be particularly relevant. London | Remote If you’re interested in owning a product at the intersection of AI, machine learning and infrastructure, I’d be happy to tell you more.

Research Scientist / Research Engineer – LLM Post-Training & Reinforcement Learning Location: San Francisco Bay Area Industry: Frontier AI / Foundation Models Compensation: Highly competitive Seniority: Multiple levels The Opportunity We're working with an ambitious frontier AI research company in the Bay Area, building next-generation foundation models and exploring new approaches to advancing model intelligence, reasoning and computational efficiency. As part of their continued growth, they're looking to hire talented Research Scientists and Research Engineers specialising in LLM Post-Training and Reinforcement Learning. You'll have the opportunity to tackle complex, open-ended research challenges, contribute to the development of novel training methodologies and play a key role in shaping the capabilities of future foundation models. What You'll Be Working On The research spans several areas of modern foundation model development, including: Post-Training & Reinforcement Learning: Developing and improving RL-based training approaches, including PPO, GRPO, RLHF, preference optimisation and reward modelling. Large-Scale RL Systems: Building and optimising distributed training infrastructure, asynchronous training pipelines and high-throughput inference and rollout systems. Reasoning & Agentic Capabilities: Exploring methods to improve complex reasoning, multi-step decision-making, tool use and autonomous agent behaviour. Evaluations & Reward Design: Creating robust evaluation methodologies, synthetic environments and verifiable reward mechanisms to drive measurable improvements in model performance. Research & Experimentation: Taking ownership of research initiatives, from identifying promising ideas and designing experiments through to training, analysing results and implementing successful approaches. What We're Looking For We're interested in researchers and engineers who combine strong technical foundations with the ability to independently investigate and solve challenging research problems. Ideally, you'll bring: Practical experience training, fine-tuning or post-training large language models. Strong knowledge of reinforcement learning algorithms, particularly PPO, GRPO, RLHF and reward modelling. Experience with distributed training, large-scale GPU infrastructure or optimising RL training and inference workloads. Proficiency in Python, PyTorch and modern deep learning frameworks. Experience developing evaluations, designing experiments and interpreting model behaviour. A track record of meaningful contributions to AI research, whether through publications, open-source projects or production-scale model development. Experience with frameworks such as verl, OpenRLHF, TRL, NeMo RL, DeepSpeed or similar distributed training technologies would be particularly relevant. Why This Opportunity? This is a chance to join a research-driven organisation working at the forefront of foundation model development, where new ideas can directly influence the models being built. You'll work alongside highly capable researchers, have significant ownership over your research direction and contribute to solving fundamental challenges in AI. The team is expanding across multiple levels and welcomes both research-focused scientists and hands-on research engineers with experience developing and scaling advanced AI systems. Interested in pushing the boundaries of LLM post-training and reinforcement learning? Apply directly to us with your resume or get in touch for a confidential conversation to learn more about the company, research and opportunities available.

Summary An innovative quantum technology company is looking for a Quantum Success Engineer to support customers using advanced quantum control hardware and software. This role combines technical troubleshooting, customer onboarding and training, and relationship management. You will help researchers integrate quantum control systems into their experiments, resolve technical challenges and ensure they get the most from the technology, while sharing customer feedback with internal engineering and product teams. Responsibilities Customer onboarding and training: Support hardware installation, setup and integration, delivering technical training both onsite and remotely. Technical support: Troubleshoot customer issues involving quantum control systems, helping experimental research teams overcome technical challenges. Customer relationships: Build strong relationships with customers, respond to concerns and support customer satisfaction and long-term retention. Customer advocacy: Gather customer feedback and communicate insights to product and engineering teams to inform improvements and future development. Cross-functional collaboration: Work with internal technical and commercial teams to resolve issues and improve the overall customer experience. Requirements A PhD in experimental physics involving microwave or radio-frequency experiments, or a Master's degree in experimental physics with 1–2 years of relevant industry experience. Familiarity with Python programming. An understanding of electronics and experimental physics; experience with quantum computing is highly advantageous. Experience with embedded systems through academic projects or professional work is desirable. Strong problem-solving skills and a proactive, customer-focused approach. Excellent communication skills, with the ability to explain technical concepts clearly and professionally. Previous industry experience and/or experience in a customer-facing technical role is advantageous. Professional fluency in English. Benefits and opportunity Hybrid working: Three days in the office and two days working remotely. Generous annual leave: 30 days of paid holiday, in addition to national holidays. Competitive compensation: Attractive base salary and a 14% employer pension contribution. Company equity: Opportunity to participate in the company's long-term success through Stock Appreciation Rights. Relocation support: Relocation package and visa sponsorship for international hires and their families. Cutting-edge technology: Work at the forefront of quantum computing, supporting researchers and experimental teams worldwide. Career development: Collaborate with physicists, engineers and developers in a growing quantum technology business.

Summary An innovative deep-tech company developing next-generation quantum computing systems is seeking a Software Manager / Senior Software Manager to lead the team responsible for the software that controls and operates its hardware. With systems already deployed with customers, the company is focused on building a reliable, production-grade software platform that integrates quantum technology with existing HPC, GPU and data-centre infrastructure. This role combines software architecture, technical leadership and people management, with ownership of the platform, roadmap and customer-facing reliability. Responsibilities Technical leadership: Own the architecture and evolution of the low-level software stack, ensuring reliable integration with compute, networking, security and data-centre infrastructure. SDK and APIs: Lead the development of developer tools and APIs for hybrid quantum-classical applications. Team management: Lead, mentor and develop software engineers, setting priorities, standards and resource plans. Roadmap and delivery: Prioritise features and upgrades in collaboration with commercial, customer-facing and engineering teams. Performance and reliability: Improve latency, throughput, uptime and recovery in collaboration with hardware and optical engineering teams. Engineering quality: Establish best practices across testing, CI/CD, code reviews, release management, observability and incident response. Customer support: Lead technical escalations and root-cause analysis, ensuring issues are resolved effectively. Cross-functional collaboration: Promote documentation and shared ownership, and communicate technical progress to senior leadership. Requirements 10+ years of experience building and shipping production software, including leading engineering teams. Experience with HPC systems and/or large-scale data-centre infrastructure. Strong Python and C++ and/or Rust skills, with solid Linux systems knowledge. Experience with hardware-integrated software, such as real-time systems, instrument control, embedded software or FPGA integration. Strong engineering practices across testing, CI/CD, releases, monitoring and root-cause analysis. Experience supporting deployed customer systems and communicating calmly under pressure. Strong problem-solving, decision-making and prioritisation skills. A degree in Computer Science, Electrical Engineering, Physics or a related quantitative discipline, or equivalent experience. Desirable: Quantum computing, quantum control systems or scientific instrumentation. Digital signal processing, data acquisition, timing or synchronisation. Experience in high-reliability sectors such as semiconductors, telecommunications, medical devices, defence or aerospace. Experience scaling deep-tech products from R&D into commercial deployment. Benefits and opportunity Work on advanced quantum computing technology with real-world customer deployments. Shape the architecture, roadmap and reliability of a growing software platform. Combine technical decision-making with team leadership and development. Collaborate across software, hardware and optical engineering. Help bring next-generation computing technology into modern AI and HPC environments. Competitive salary and the opportunity to make a meaningful impact in an ambitious deep-tech business.

Research Scientist / Engineer, Inference San Francisco | Permanent DeepRec has partnered with a frontier AI research lab that designs its own model architectures from the ground up. Serving cost is treated as a research question here, so inference sits at the centre of the work rather than the end of it. The Opportunity You'll own latency and throughput across two serving stacks. One handles production traffic for customers. The other powers test-time compute and RL rollouts for the research team, so every gain you make speeds up the research itself.   What You'll Do Cut latency and raise throughput for model serving at very high concurrency Speed up RL rollout generation and scale test-time compute for research Tune batching, caching, load balancing and parallelism across the serving stack Write and profile GPU kernels in CUDA and Triton Apply quantization, distillation, speculative decoding and low-precision numerics Shape model and hardware co-design with serving cost in view from day one Own benchmarking, testing and reliability for every inference service What You'll Bring Essential Production experience with vLLM, TensorRT-LLM or similar at high concurrency Hands-on continuous batching and KV cache management Quantization for inference in production Profiling and performance analysis with CUDA and Triton Desirable Speculative decoding or multi-node serving Performance tuning on H100 or B200 hardware Kubernetes and production infrastructure We encourage you to apply even if you don't meet every requirement. The right mindset matters as much as the right CV.     Apply Now If you've spent your career making models fast, this is a place where that work shapes the research agenda. Send your CV and we'll share the full brief in a confidential conversation