Research

Machine Learning Researcher, Audio $250,000 – 300,000+, Equity + Bonus Remote (US & Europe) / San Francisco, CA (Hybrid preferred) Full-time / PermanentDeepRec has partnered with a fast-growing, revenue-generating voice AI company empowering enterprises to build AI phone agents at scale. Recent Series C funding with backing from leading Silicon Valley investors, they are building the models and infrastructure that make voice the primary interface between businesses and their customers.This company has built all of their current models completely in-house, and every model ships to real, paying customers almost immediately. No speculative research track here. If you want your work to hit production within weeks, not years, this is that role. The OpportunityThe research team are working toward a single, ambitious goal: a fully speech-to-speech conversational AI model that understands and responds like a human, in real time. You'll work across the core building blocks of that roadmap, such as: speech-to-text, text-to-speech, neural audio codecs, and getting LLMs to understand and reason over audio directly.You'll take ideas from theory through large-scale training to production inference serving millions of calls a day, working closely with engineering and product teams to get your research into real customer environments fast. What You'll DoBuild TTS models that sound natural, expressive, and humanBuild STT systems that stay accurate with accents, background noise, and messy phone linesWork on neural audio codecs, compressing audio efficiently without losing qualityExplore LLM-Audio UnderstandingPrepare and manage large audio datasets, and design how models are trained on themRun training across many GPUs at once, keeping an eye on cost and speedRun fast, well-designed experiments to test what actually worksMake sure models run fast and reliably in production, not just in benchmarksWhat You'll Bring EssentialA genuine, self-driven interest in this area of research, shown through your own projects or papers, not just what a past employer asked you to doHands-on experience building or improving TTS, ASR, speech-to-speech, or neural audio codec systemsA track record of original, hands-on technical work, rather than off-the-shelf tools or common tutorial-style projectsStrong Python skills and experience training or running large modelsComfortable working autonomously DesirableExperience getting LLMs to understand or reason about audio - the team's top priority right nowExperience with model distillationExperience fine-tuning or training language or speech models with reinforcement learningPublished research or open-source work in speech or language AIBackground working with real-time speech systems or phone-based systemsA PhD is welcome but not required. Strong, independent work matters more than qualificationsWe encourage you to apply even if you don't meet every requirement. The right mindset and genuine curiosity matter as much as the resume. What's In It For YouGround-floor seat in a lean and growing research team with real scope to shape it as it scalesEvery research output ships to real customers, no long speculative research projectsHigh autonomy to shape your own research direction, tooling, and (for senior hires) the team itselfWork across ASR, TTS, neural codecs, and the frontier of LLM-audio understanding & speech-to-speech modellingWell-capitalised, fast-moving environment without big-lab bureaucracyHealthcare, dental, vision, meaningful equity, and every tool you need to succeedRemote-friendly across the US

Senior Reinforcement Learning EngineerZurich | Hybrid | Full-timeYou’ve already deployed reinforcement learning on real robots. Now you can apply that experience to autonomous excavators working across different machines, sites and soil conditions. You’ll join a Series A robotics company taking Physical AI into construction, with systems already deployed across multiple countries.You’ll build learning-based planning and control systems that work outside the simulator. That means improving simulation and sim-to-real transfer, designing data pipelines for real-world training, running experiments on physical machines and understanding why behaviour changes when conditions get messy. You’ll also integrate learned components into the wider autonomy stack and help shape how the system moves from prototype into a reliable product.This is a hands-on engineering role for someone with 2–5 years of industry experience in reinforcement learning for control or planning, who has actually deployed systems on physical robots. You’ll need strong Python and PyTorch skills, good C++, experience with GPU-accelerated simulation, and the ability to debug real-world robotic behaviour. Experience with hydraulic machinery, large-scale deployments, imitation learning or production rollout strategies would be useful.You’ll have genuine scope to influence the technical direction as the autonomy team builds a long-lived system designed to operate across the construction industry. If you want your RL work to move from simulation into machines doing real work, this is an opportunity to do exactly that.You’ll need:2–5 years’ industry RL experience in control or planningProven deployment on physical robotsStrong Python/PyTorch and good C++Experience with simulation and sim-to-realWillingness to travel when projects require itIf the challenge fits your background, let’s have a conversation about the role and the problems you’d be working on.

DeepRec.ai is supporting a high-growth cybersecurity startup building security solutions for the AI era, focused on protecting AI applications, agents, identity infrastructure, and sensitive enterprise data. (Fully remote position) We're looking for a Threat Researcher with a strong offensive security background to research emerging attack techniques across AI and enterprise environments.   What You'll DoResearch attacks against LLMs, AI agents, APIs, identity systems, and data flowsInvestigate prompt injection, agent manipulation, token abuse, privilege escalation, and data exfiltrationDevelop threat models, attack simulations, and working PoCsBuild research tooling and test environments using Python, Docker, and cloud platformsCollaborate with engineering and product teams to turn research into defensive capabilitiesPublish original research through blogs, talks, advisories, or open-source toolingApply frameworks such as MITRE ATT&CK to emerging AI attack techniquesWhat We're Looking For6–10 years in threat research, red teaming, offensive security, or security engineeringStrong hands-on offensive security and vulnerability research experienceDeep understanding of AI/LLM and agentic architecturesStrong knowledge of IAM, OAuth/OIDC, tokens, privileges, and DLPStrong Python and cloud/container experienceTrack record of publicly shared security research, tooling, or technical writingAbility to communicate complex security research clearly to technical and non-technical audiencesThis is a highly hands-on research role with significant ownership over the research agenda and the opportunity to work on new attack surfaces emerging from increasingly autonomous AI systems.

Principal AI Security Researcher Our client is a fast-growing cybersecurity company building security infrastructure for AI systems. They're hiring a Principal AI Security Researcher to lead and scale their AI security research and red-teaming function. Highly technical leadership role. Candidate should have spent significant time attacking, evaluating, and securing modern AI systems. Sits at the intersection of offensive security, adversarial AI research, and applied engineering. What you'll be doingLead and grow teams focused on AI red teaming, adversarial research, and AI security engineeringDesign and execute advanced attacks against LLMs, GenAI applications, AI agents, and multi-agent systemsResearch prompt injection, jailbreaks, indirect prompt injection, tool abuse, agent manipulation, data exfiltration, model misuse, adversarial behaviorBuild automated systems for continuous AI security testing and adversarial evaluationEstablish methodologies and infrastructure for testing AI systems at scaleLead research into emerging attack surfaces across agent orchestration, tool use, and multi-agent protocolsTranslate research into guardrails, detection mechanisms, monitoring, and security controlsPartner with engineering and product to build security into AI systems through the development lifecycleDefine AI security testing frameworks informed by OWASP, MITRE ATLAS, NIST AI RMFShape technical roadmap and long-term research strategy for the company's AI security platformRepresent the company externally through research, publications, conferencesWhat we're looking forSignificant experience leading security research, offensive security, AI security, or R&D teamsDeep hands-on expertise in AI red teaming, adversarial ML, LLM security, or GenAI securityStrong understanding of how modern AI systems are built, attacked, evaluated, deployedPractical experience researching or exploiting vulnerabilities in LLMs, AI applications, or agentic systemsExperience building security testing frameworks, offensive tooling, automated evaluations, adversarial testing infrastructureComfortable operating at strategic and technical level, close enough to the research to challenge assumptionsTrack record building high-performing technical teams and taking research from concept to implementationStrong communication across engineering, product, security, and executive stakeholders5+ years across cybersecurity, AI security, ML security, adversarial research, or related fieldParticularly interesting backgroundsPublished AI security or adversarial ML researchConference presentations (Black Hat, DEF CON, OWASP events)Open-source AI security contributions, standards, or research communitiesBuilt automated red-teaming or adversarial evaluation platformsDeveloped AI guardrails, runtime security systems, anomaly detection, AI monitoring infrastructureHands-on experience with agentic frameworks and multi-agent orchestrationExperience securing AI workloads across major cloud environmentsBackground in AI trust & safety, adversarial ML, offensive security, or AI researchExperience evaluating frontier models or complex AI/agentic systems in production