Lead Evaluation Engineer — Speech & Multimodal Models

How do you measure whether an AI voice truly sounds real — and prove it with data?

You’ll join an AI team developing large-scale speech and multimodal systems for real-time interaction — models that generate, clone, and understand voice with natural expression and precision.

This is a founding evaluation role, in a new dedicated Evals team defining how these models are measured, improved, and deployed safely at scale. You’ll design objective and subjective evaluation pipelines, run large-scale human studies, and build automated systems that turn perception into measurable signal.

Your work will span every stage of model development — from research to production — collaborating with speech, audio, and ML teams to close the loop between modelling, feedback, and user experience.

What you’ll do:
• Build and scale evaluation pipelines for TTS, voice conversion, and ASR systems
• Design human studies for subjective testing (e.g. MOS, ABX)
• Define and implement objective metrics (WER, intelligibility, naturalness, prosody)
• Automate evaluation dashboards and reporting systems
• Train auxiliary models to capture new evaluation dimensions
• Collaborate across data, model, and product teams to drive measurable improvement
• Establish and scale the evaluation function as the team grows

You’ll bring:
• Strong experience building or running eval systems for speech or multimodal models
• Familiarity with ASR, TTS, or voice cloning pipelines
• Experience designing user studies or subjective model evaluation
• Solid understanding of statistics and experimental design
• Proficiency in Python and ML frameworks (PyTorch, Hugging Face, etc.)
• Strong communication skills and cross-functional mindset

Why this role:
This is a rare chance to build the evaluation foundation for models already deployed globally — shaping how next-generation speech systems are measured and improved. You’ll have the autonomy to define standards, lead future hires, and see your work directly impact millions of real-world interactions.

Fully remote (EU timezones preferrred), global team. Competitive salary + meaningful stock options.

The company are well funded, with a 9 figure funding round and significant runway for meaningful growth, lots of compute and hiring! 

Apply today. Everyone will get a response.

Teach AI how to reason — safely, transparently, and at scale.

How do we move beyond pattern-matching into true machine reasoning? This Applied Scientist role puts you at the centre of that challenge — developing models that can reason, explain their logic, and make verifiable decisions across complex, high-stakes industries.

You’ll join a well-funded startup building domain-specific reasoning systems and agentic AI for sectors like medtech, aerospace, advanced manufacturing  — where reliability and interpretability aren’t optional.

Your work will focus on post-training large multimodal models, applying the latest techniques in RLHF, DPO, and preference learning to make AI systems more consistent, factual, and aligned with human reasoning. You’ll design the frameworks that turn raw model potential into transparent, trustworthy intelligence.

You’ll develop and optimise post-training pipelines, implement reward modelling for reasoning depth and factual accuracy, and build evaluation frameworks for verifiable, human-aligned behaviour. Working with proprietary and synthetic datasets, you’ll run end-to-end experiments and deploy your methods directly into production.

You’ll bring a background in transformer-based model training (LLM, VLM, MLLM), post-training or alignment (RLHF, DPO, reward modelling), and strong practical skills in Python and PyTorch. Curiosity about reasoning agents, hybrid learning, and interpretability research will help you thrive here.

Bonus points for experience in multimodal reasoning, evaluation and verification, or prior research contributions in alignment or reasoning systems.

The company has raised $20M+ (Series A announcement imminent) and already partners with Fortune 100 and 500 customers. Founded by an entrepreneur with a prior billion-dollar exit, the AI team alone is scaling from 11 to 40+ this year.

Comp: $200K–$320K base (negotiable depending on experience) + bonus + stock + benefits
Location: SF Bay Area (remote for now; hybrid later in 2026)

If you’re excited about defining how AI systems reason, decide, and explain themselves — we’d love to hear from you.

All applicants receive a response.

Looking to push the boundaries of generative AI for real-time interaction?

You'll be joining a well-funded startup working on Multimodal AI where voice, vision, and language come together. 

They're building generative models for natural conversational experiences that need to perform in real-time.

Your mission

You'll be building and optimising diffusion or flow-matching models that power their speech and audio generation. 

This means developing production-ready architectures that can generate controllable, high-quality output at scale.

You'll own the full research-to-production pipeline - from architecture design and training through deployment and optimisation. 

Your work will directly impact how millions of AI characters sound and interact.

Your focus

• Design and train large-scale diffusion or flow-matching models
• Develop novel architectures and training techniques to improve controllability and quality
• Build evaluation systems to measure generation quality and model behaviour
• Work from low-level performance optimisations to high-level model design

What you'll bring

• Proven track record building diffusion models or flow-matching systems
• Experience training large models (3B+ parameters) with distributed systems

Nice to have

• Experience with audio or speech generation
• Publications or open-source contributions in diffusion models or generative AI

Remote in Europe with competitive comp + stock.

Want to own complex 3D vision systems that operate outside controlled conditions?

This role is for someone who enjoys building perception systems that have to perform reliably on real data, under real constraints. 

You’ll join a small engineering group where senior ICs take responsibility for designing, evolving, and maintaining core 3D vision capabilities end-to-end.

There’s very little process overhead. Engineers are trusted to define solutions, make decisions, and move work forward quickly. The problems are open-ended, technically demanding, and require strong judgement as much as technical depth.

What you’ll do:

• Own and evolve key parts of a 3D vision pipeline from early design through to deployment
• Build and refine geometry-heavy vision systems, integrating learned models where appropriate
• Prototype ideas quickly and evaluate them against realistic, imperfect data
• Optimise models and pipelines for accuracy, performance, and robustness
• Push successful approaches into production-quality implementations
• Work closely with adjacent engineers while independently driving complex technical problems

What you’ll bring:

• Experience building real 3D vision systems, not just experimenting with libraries
• Strong Python and/or C++ skills, with care for performance and reliability
• Comfort working across geometry, optimisation, and modern deep learning approaches
• Experience tuning systems over time and making pragmatic technical trade-offs
• Ability to operate effectively without detailed specifications or rigid roadmaps
• A sharp, curious mindset and the ability to learn quickly in demanding environments

Why people join:

You’ll work alongside a small, elite group of engineers who take their craft seriously and value depth, speed, and ownership. 
The bar is high, the scope is meaningful, and good work is recognised quickly. If you enjoy being trusted with important systems and want to work with peers who are equally strong technically, this environment supports that.

Package

• Compensation: $200,000 – $500,000 base (negotiable d.o.e) + equity
• Benefits: Medical, dental, vision, 401(k)
• Location: San Francisco (on-site)
• Employment: Full-time

If this sounds like work that interests you, and you want to be working at the cutting edge of technology with an elite team, please apply now!

All applicants will receive a response.

Build the 3D perception that gives AI agents real spatial intelligence.

How do AI systems truly see and reason about 3D geometry? This Applied Scientist role puts you at the centre of that challenge — developing models that bridge the physical world and intelligent reasoning systems.

You’ll join a well-funded startup building AI agents for advanced design and engineering workflows — across manufacturing, aerospace, and medtech. Your work will enable agents to understand CAD data, meshes, and point clouds deeply enough to plan, analyse, and make autonomous decisions.

This is a rare opportunity to establish the 3D foundation within the research team. You’ll define evaluation strategies, model objectives, and technical direction — building models that become the perception backbone for intelligent agents.

What you’ll do:
• Develop models that learn transferable 3D representations across CAD, mesh, and point cloud data
• Handle messy, lossy, real-world data — not just clean synthetic geometry
• Scale training across segmentation, classification, correspondence, and eventually generation
• Design robust evaluation pipelines for continuous performance monitoring
• Work toward a unified 3D foundation model supporting both discriminative and generative tasks

You’ll bring:
• Deep expertise in 3D computer vision (PhD or equivalent experience)
• Strong knowledge of modern 3D architectures (PointNet++, MeshCNN, Gaussian Splatting, Diffusion, VLMs)
• Proven ability training large-scale models in PyTorch
• Strong applied research instincts — turning papers into working systems
• Experience with multimodal or vision-language models

Bonus points:
• Background with CAD data or industrial design workflows
• Experience in robotics, autonomous driving, or AR/VR 3D perception
• Familiarity with SLAM, pose estimation, or differentiable rendering

You’ll join a small, research-driven team with full autonomy and major compute access — free to explore foundational methods while delivering practical impact.

Compensation & location:
• Base salary: $200K–$300K (negotiable by level)
• Up to 20% bonus + stock
• Full medical, dental, and vision coverage
• 401k (3% match) and 20+ vacation days

Based in the SF Bay Area (currently remote, moving hybrid soon).
Applicants must hold valid US work authorisation (US Citizen or Green Card).

If you’re excited about building the 3D understanding that will power the next generation of intelligent agents — we’d love to hear from you.
All applicants will receive a response.

Want to build the large-scale RL environments frontier labs use to train agents that can truly reason and act?

This team are creating complex reinforcement learning environments — simulations where advanced agents learn to plan, adapt, and solve multi-step problems that stretch beyond standard benchmarks. The focus isn’t on training the models themselves, but on building the worlds that make meaningful learning and evaluation possible — the foundation for more capable, aligned systems.

You’ll work end-to-end across environment design, reward dynamics, and scalable simulation — developing the feedback loops that define what “good” looks like for intelligent behaviour. It’s open-ended, research-driven work where the task definition, data, and reward structure are often the hardest and most important problems to solve.

You’ll collaborate closely with researchers tackling unsolved challenges in reinforcement learning and agent behaviour, shaping experiments, scaling infrastructure, and refining how agents learn in the loop.

It suits someone with strong ML and RL experience, deep intuition for agent dynamics, and the curiosity to explore problems that don’t come with clear instructions.

On-site in San Francisco. Compensation up to $300 K base (negotiable, depending on experience) plus equity.

If you want to help build the environments that teach the next generation of AI systems how to think, act, and adapt — we’d love to hear from you.

All applicants will receive a response.

Interested in advancing how agents and LLMs learn from feedback in realistic environments?

You’ll be joining a research-driven AI company building reinforcement learning simulation environments for agents and large language models, with a focus on post-training, evaluation, and scalable supervision. 

Their tools are already used in production by leading AI labs and enterprises and due to demand they are growing fast.

As a Research Scientist, you’ll work hands-on on fundamental problems spanning LLM post-training, RL environments, and agentic evaluation. Your work will shape core methods and benchmarks, and you’ll see your research deployed into production systems. The team actively publishes and collaborates with external research labs, with recent work appearing at ACL and NeurIPS.

What you’ll do

• Conduct research on LLM post-training methods (RLHF, RLAIF, RLVR)

• Design and build realistic RL simulation environments for agents

• Develop agentic evaluation and supervision frameworks

• Create and maintain benchmarks for emerging AI capabilities

• Collaborate with engineers to take research from idea to deployed systems

What you’ll bring

• Experience in applied research in reinforcement learning, LLM post-training, or agent-based systems

• Strong understanding of transformer architectures and LLM fine-tuning

• Ability to translate research ideas into working, production-ready systems

Nice to have

• Publications at top-tier venues (NeurIPS, ICML, ACL, EMNLP)

• Experience working on evaluation, safety, or oversight for advanced AI systems

• Prior work on large-scale training or simulation environments

SF-based. Compensation up to $300k base (flexible, DOE) plus equity, unlimited PTO, and benefits.

Interested in working on the foundations of AI training, evaluation, and safety—while publishing high-quality research that ships?

All applications will receive a response.

Want to build speech AI that actually sounds human?

You'll be joining a well-funded speech AI startup with strong customer traction. They're building ultra-realistic voice technology that handles natural laughter, breathing, seamless language switching, and accurate pronunciation across languages and accents.

As a Staff Research Engineer, you'll work hands-on to expand their foundation models and push the boundaries of what's possible in speech AI: exploring multilingual capabilities, long-context generation, full-duplex modeling for natural conversations with interruptions, and novel architectures that balance speed with control.

What you'll do

• Conduct research to advance their core speech models and extend product capabilities
• Develop and experiment with new model architectures and training approaches
• Work on large-scale model training and data systems
• Collaborate with the team to take research from concept to deployed systems

What you'll bring

• 3+ years of experience in speech synthesis, audio generation, or generative modeling
• Experience with audio generation using LLMs
• Solid background in modern language model architectures
• Proven ability to ship research into production systems
• Experience training large-scale models

Nice to have

• Published research in speech or generative modeling
• Experience with real-time speech systems or multimodal models

Ideally in SF, but can also consider remote worldwide. Comp is up to $250K base DOE, plus equity.

Want to define how AI generates coherent video over minutes, not seconds?

This role sits at the heart of one of the hardest open problems in generative media: long-form video generation!

You’ll join a small, research-driven team building a multi-modal foundation model that reasons jointly across image, text, and audio. Their work powers a creative platform used to generate controllable, expressive video - and the underlying model is already in production.

As a Research Scientist focused on long video generation, you’ll work on the architectural problems that emerge once sequences stop being toy-length.

You’ll spend your time pushing sequence models to handle multi-minute videos without collapse. 

What you’ll work on

• Architectures for long-form, auto-regressive video generation
• Causal attention and long-context modelling strategies
• Techniques for temporal and semantic coherence over extended sequences
• Memory-efficient transformers and sequence compression
• Translating research into production-grade pipelines
• Publishing and presenting work internally and externally

You’ll fit well here if you’re comfortable operating at the intersection of theory and systems. 

Someone who can read the latest long-context papers, prototype quickly in PyTorch, and reason about what scales when models move from experiments to real users.

The team works fully in-person in San Francisco or New York.

What you’ll bring

• PhD or equivalent research/industry experience in ML or sequence modelling
• Deep understanding of transformers, attention, and auto-regressive generation
• Experience with long-context or memory-efficient modelling
• Strong Python and PyTorch skills
• Evidence of real research impact or large-scale deployment

Package

• Salary: Negotiable depending on experience
• Meaningful equity
• Medical, dental, and vision cover
• 401(k)
• Lunch and snacks provided
• Fully in-person role (SF or NYC)

If you want to work on long-form video generation problems, this is one of the few places doing it properly - please apply now!

All applicants will receive a response.

Looking to define ASR strategy for the next generation of social AI?

You'll be joining a well-funded social AI company building lifelike AI characters that interact naturally across voice, video, and text. Founded by a prominent tech entrepreneur, they're creating new media formats for AI-driven interaction where agents handle group conversations, interruptions, and multi-agent dynamics.

Your mission

You'll own the ASR function from day one - starting with evaluating and implementing existing solutions, then moving toward building proprietary models as the platform scales. This means hands-on work testing APIs and open-source models, followed by developing custom systems for multi-agent group conversations and social interactions.

You'll shape the technical direction, balance short-term delivery with long-term innovation, and drive individual research initiatives while collaborating on broader team objectives.

Your focus

• Define and execute the ASR roadmap from evaluation through production deployment
• Build and train models that handle natural conversation dynamics
• Develop evaluation systems to measure accuracy, speed, and reliability
• Define data requirements and create pipelines for ASR training
• Work from low-level performance optimizations to high-level architecture decisions

What you'll bring

• Proven track record building and deploying ASR systems at scale
• Strong familiarity with SOTA ASR models and architectures (Whisper, Conformer, etc.)
• Understanding of data quality assessment for speech systems

Nice to have

• Experience leading technical initiatives or ML teams

Remote with competitive comp + stock.

Ready to define the future of social AI interactions? Apply today.