--- title: 'Applied AI Researcher (Dublin, CA) at Articul8' canonical: 'https://feeny.ai/job/applied-ai-researcher-dublin-ca-articul8-dublin-ffzj33rv5pnm' type: 'job' last_seen: '2026-09-10' --- # Applied AI Researcher (Dublin, CA) at Articul8 - **Company:** Articul8 - **Location:** Dublin, Ireland - **Employment:** full-time - **Posted:** 2026-05-05 - **Last confirmed live:** 2026-09-10 - **Apply:** https://jobs.ashbyhq.com/articul8/3ff51e5b-fc0c-4dcd-929e-4d5463cfe467/application **Skills:** Python, Git, Probability, Optimization, Linear Algebra, NLP, Computer Vision, Reinforcement Learning, Information Retrieval, Distributed Computation, Experiment Tracking, Python Libraries, PyTorch DDP, DeepSpeed, FSDP, RLHF, DPO, Reward Modeling, AWS, GCP > The Applied AI Researcher will design and run experiments, build training and evaluation pipelines, and ship research into production for a domain-specific GenAI platform. Responsibilities include model training across LLMs, reinforcement learning, and multimodal understanding while developing agentic research infra... ## Job description About Us: Articul8 was born from a simple belief: GenAI should work for the enterprise, not the other way around. Our platform — combining domain-specific models, autonomous agentic reasoning (ModelMesh™), reliable model evaluation (LLM-IQ™), and multimodal understanding — serves regulated industries such energy, semiconductor, finance, aerospace, supply chain, and more. Trusted by Fortune 500 enterprises, we bring together research, engineering, product, and domain expertise to deliver AI that meets the accuracy, explainability, and auditability standards that high-stakes environments demand. Job Description: Articul8 AI is seeking an Applied AI Researcher to advance our domain-specific GenAI platform. You will design and run experiments, build training and evaluation pipelines, and ship research into production. This role spans model training, reinforcement learning, multimodal understanding, and knowledge representation. Responsibilities: - Architect and orchestrate massively parallel AI research workflows — design experiments that leverage fleets of agentic AI systems to explore hypothesis spaces, hyperparameter landscapes, and architectural variations at a scale and speed no single researcher could achieve alone - Design, train, and iterate on models across the full GenAI stack — LLMs, VLMs, embedding models, rerankers, and reward models — using agentic pipelines that autonomously manage data preprocessing, training runs, evaluation sweeps, and result synthesis - Go deep: push the frontier of domain-specific AI — conduct rigorous, first-principles research into model architectures, training dynamics, reinforcement learning, and knowledge representation, using AI agents to accelerate literature review, ablation studies, and mathematical analysis - Go broad: span disciplines and modalities — amplify your expertise across NLP, computer vision, multimodal understanding, agentic reasoning, and domain science by delegating exploration, prototyping, and benchmarking to parallel agent systems so you can synthesize insights across fields simultaneously - Build agentic research infrastructure — develop and contribute to shared tooling, libraries, and platforms that enable every researcher on the team to orchestrate autonomous experiment pipelines, data processing workflows, and evaluation harnesses at scale - Ship research into production at velocity — collaborate with engineering, product, and domain experts to integrate breakthroughs into the platform rapidly, using agentic CI/CD and automated integration testing to compress the research-to-deployment cycle - Amplify collective intelligence — document findings, publish at top-tier venues, and build internal knowledge systems that agentic tools can index and reason over — turning every insight into a force multiplier for the entire team - Continuously raise the ceiling on human potential — proactively identify bottlenecks in your own workflow and the team's, then design or adopt efficient, scalable solutions that eliminate them — treating your own augmentation as a core research output Required Qualifications: - Education: PhD in Computer Science, Machine Learning, or a related field; or MSc with 4+ years of post-graduation research experience. - Model development: You have trained or fine-tuned at least one neural model end-to-end — data preparation through evaluation. You understand why your model converges or doesn't, not just how to launch a training run. - Technical foundations: Strong working knowledge of probability, optimization, and linear algebra applied to at least one of: NLP, computer vision, reinforcement learning, or information retrieval. You can derive the math behind the methods you use. - Infrastructure: Experience building training or evaluation pipelines that handle real data — preprocessing, distributed computation, experiment tracking, and reproducibility. - Software engineering: Production-quality Python. You write code others can read, test, and extend. Fluent with Git and collaborative development workflows. Preferred Qualifications: - Experience with distributed training frameworks (PyTorch DDP, DeepSpeed, FSDP) — you understand gradient synchronization and can debug multi-GPU failures. - Published at NeurIPS, ICML, ICLR, ACL, EMNLP, CVPR, or equivalent. Quality of contribution matters more than count. - Hands-on experience with post-training methods (RLHF, DPO, reward modeling) — beyond reading papers. - Practical cloud infrastructure experience (AWS, GCP, or Azure) for ML workloads — you can provision resources, manage jobs, and troubleshoot training failures. Professional Attributes (Code42): - Practice Humility: You ask questions even when you think you know the answer. You seek feedback early, learn from anyone regardless of title, and treat every experiment — especially the failures — as data. - Bias for Outcomes: You measure your work by what changed, not what you tried. You ship results, not slide decks. When a deadline is real, you find a way. - Care Deeply: You treat every problem as yours to solve. You review your own work with the rigor you'd want from a reviewer. You help teammates without being asked. - Dare to Do the Impossible & Embrace Scarcity: You set goals that make you uncomfortable. When told something can't be done, you find a way or a better question. Constraints sharpen your thinking, not slow it down. - Build a Better World: You believe AI should make things meaningfully better for real people. You hold yourself accountable not just for whether your model works, but for what it does in the world. If you're ready to join a team that's changing the game, apply now to become a part of the Articul8 team. ## About Articul8 ## Company Overview - **One-liner**: Articul8 provides a full-stack domain-specific generative AI platform for enterprises in regulated industries, enabling them to build expert-level AI applications within their own IT environments. - **Entity Type**: Private (Series B) - **Headquarters**: Dublin, California, USA - **Founded**: 2024 - **Founders**: Arun K. Subramaniyan ## Core Business - **Primary industry/industries**: Enterprise AI, Generative AI, Domain-Specific AI - **Target customers**: Large enterprises in regulated industries including energy, manufacturing, aerospace, semiconductors, and financial services (B2B, Enterprise) - **Mission or purpose statement**: To transform enterprise data and expertise into powerful engines of growth, value, and impact by making it straightforward for companies to build sophisticated, enterprise-scale, and expert-level GenAI applications that encode their domain expertise. ## Products & Services - **Articul8 Platform**: A full-stack GenAI platform that processes enterprise data into AI applications, facilitating decision-making, data automation, and adherence to security and privacy standards. It features autonomous agentic reasoning, model evaluation and dynamic routing (LLM-IQ™), and a proprietary ModelMesh™ for multi-agent collaboration. - **Domain-Specific Models (DSMs)**: - **A8-Energy**: Developed with EPRI; trained on 10,000+ specialized energy datasets for expert reasoning in the energy sector. - **A8-SupplyChain**: Optimized for manufacturing and supply chain operations; reasons over complex technical documentation without data replication. - **A8-Fin**: Finance-focused DSM for tasks such as tabular analysis, portfolio management, and compliance. - **A8-Semicon**: Verilog-capable DSM for semiconductor engineering, integrating domain knowledge with reasoning for complex chip design workflows. - **Hyper-Personalized Agent Models**: Developing hyper-personalized agents for enterprise users. ## Market Standing - **Valuation/Market Cap**: $500 million pre-money valuation (Series B, January 2026) - **Key Metric**: Total funding of $35M raised (Series B, first tranche); $100 million in total contract value; projected annual recurring revenue of just over $57 million for 2026 - **Notable Investors/Partners**: Adara Ventures (lead), Aditya Birla Ventures, Accel, Peak XV Partners, NXC, and 14+ others. Partners include Nvidia, Google Cloud, AWS, and Databricks. Customers include Hitachi Energy, AWS, Franklin Templeton, Intel, AIAA, Itochu Techno-Solutions Corporation, Uptycs, and NIQ. - **Growth Signals**: Revenue-positive with 29 paying customers; 5x valuation increase from Series A ($100M post-money) to Series B ($500M pre-money); expanding internationally with focus on Europe and Asia (Japan, South Korea, India); recognized by Gartner as a Tech Innovator in Domain-Specific AI for Manufacturing and Energy. ## Competitive Advantages - **Domain-Specific Focus**: Unlike general-purpose models, Articul8’s DSMs are purpose-built for regulated industries, achieving 90%+ accuracy versus ~60% for competing models in proprietary benchmarks, and 28% more accurate than leading general LLMs. - **Enterprise-Grade Security**: SOC 2 Type II compliant with certified security and confidentiality standards, observability, auditability, and traceability at every step. - **Cost Efficiency**: 3.5x cheaper than best open-source models, with a 2x performance boost over the latest open-source state-of-the-art models. - **Proven Performance**: Matches or exceeds proprietary models like Google Flash 2.0 and GPT-4o at a fraction of the compute cost. - **Customer Lock-In**: Deep integration into customer IT environments and domain-specific workflows creates high switching costs. ## Strategic Focus - **International Expansion**: Scaling operations in Europe (backed by Adara Ventures and the European Investment Fund) and Asia (Japan, South Korea, India). - **Heritage AI Venture**: Raising $30-50 million for an India-based AI entity focused on traditional knowledge systems (Sanskrit, classical Indian languages) with applications in drug discovery, metallurgy, healthcare, and education. - **Product Development**: Expanding research and product development, particularly in agentic reasoning systems and hyper-personalized models. - **Marketplace Availability**: Available on AWS, Microsoft Azure, Google Cloud Platform, and Databricks to simplify deployment and integration. ## Why Work Here - **Culture**: 80% of the 75-person team is focused on R&D, indicating a strong engineering and research culture. The company is revenue-positive and not cash-strapped, providing stability. - **Remote/Hybrid Policy**: Offers remote and hybrid roles across offices in Dublin, CA (USA), Brazil, and India. - **Notable Perks**: Opportunity to work on cutting-edge domain-specific AI for regulated industries; involvement in a high-growth spinout from Intel with strong backing from top-tier VCs; chance to contribute to heritage AI initiatives preserving ancient knowledge systems. - **Engineering Culture**: Emphasis on autonomous decisions and actions, automated data intelligence, and building expert-level GenAI applications. Roles include Applied AI Researchers, Software Engineers, and Infrastructure Engineers. ## Sources 1. [articul8.ai](https://www.articul8.ai/) 2. [cbinsights.com](https://www.cbinsights.com/company/articul8-ai) 3. [techcrunch.com](https://techcrunch.com/2026/01/07/intel-spin-off-articul8-is-halfway-to-70m-ai-funding-round-at-500m-valuation/) 4. [builtin.com](https://builtin.com/company/articul8-ai) 5. [pitchbook.com](https://pitchbook.com/profiles/company/550870-03) ## Other roles at Articul8 - [Builder Product Manager, AI Platform & Agentic Workflows](https://feeny.ai/job/builder-product-manager-ai-platform-agentic-workflows-articul8-dublin-26fm156mng11) — Dublin, Ireland - [Machine Learning Engineer](https://feeny.ai/job/machine-learning-engineer-articul8-dublin-hj5recg6htpe) — Dublin, Ireland - [Infrastructure Engineer (India)](https://feeny.ai/job/infrastructure-engineer-india-articul8-brazil-wadq2s02sb00) — Brazil - [Infrastructure Engineer (Brazil)](https://feeny.ai/job/infrastructure-engineer-brazil-articul8-brazil-17gjyty3g16w) — Brazil - [Software Engineer (India)](https://feeny.ai/job/software-engineer-india-articul8-brazil-xgkd5dbqes6z) — Brazil - [Software Engineer (Brazil)](https://feeny.ai/job/software-engineer-brazil-articul8-brazil-07c8ctsqw9qp) — Brazil - [Principal Applied AI Researcher - Domain- Specific Models (Dublin, CA)](https://feeny.ai/job/principal-applied-ai-researcher-domain-specific-models-dublin-ca-articul8-dublin-28ps4ypf0bha) — Dublin, Ireland - [Principal Applied AI Researcher - Domain- Specific Models (India)](https://feeny.ai/job/principal-applied-ai-researcher-domain-specific-models-india-articul8-bengaluru-vwzseddga10j) — Bengaluru, India - [Principal Applied AI Researcher - Domain- Specific Models (Brazil)](https://feeny.ai/job/principal-applied-ai-researcher-domain-specific-models-brazil-articul8-brazil-fqjet2zxmabm) — Brazil - [Staff Applied AI Researcher - Agentic Reasoning Systems (India)](https://feeny.ai/job/staff-applied-ai-researcher-agentic-reasoning-systems-india-articul8-dublin-gfx6w6hzbkfg) — Dublin, Ireland