Forward Deployed ML Engineer, Agents

About AION AION is building an interoperable AI cloud platform by transforming the future of high-performance computing (HPC) through its decentralized AI cloud. Purpose-built for bare-metal performance, AION democratizes access to compute and provides managed services, aiming to be an end-to-end AI lifecycle platform—taking organizations from data to deployed models using its forward-deployed engineering approach. AI is transforming every business around the world, and the demand for compute is surging like never before. AION thrives to be the gateway for dynamic compute workloads by building integration bridges with diverse data centers around the world and re-inventing the compute stack via its state-of-the-art serverless technology. We stand at the crossroads where enterprises are finding it hard to balance AI adoption with security. At AION, we take enterprise security and compliance very seriously and are re-thinking every piece of infrastructure from hardware and network packets to API interfaces. Led by high-pedigree founders with previous exits, AION is well-funded by major VCs with strategic global partnerships. Headquartered in the US with global presence, the company is building its initial core team in India/UK. Who You Are You're a hands-on AI engineer with 3-5+ years of experience building production-grade multimodal AI systems and LLM applications. Your responsibilities mirror those of a hands-on AI startup CTO—you work in small teams to own delivery of high-stakes customer projects, embedding directly at client sites to architect, build, and deploy intelligent agent solutions. You're equally comfortable writing production code, presenting technical solutions to C-level executives, and debugging complex AI systems on factory floors or in customer data centers. You've shipped voice agents, video processing systems, or conversational AI to production. You thrive translating ambiguous business requirements into concrete technical solutions that create measurable impact. You're comfortable working across the full AI deployment lifecycle—from use case discovery and solution architecture to multimodal agent development, MLOps pipeline implementation, and production optimization. You understand what makes agents perform well in production and how to systematically improve quality through observability and evaluation. Experience with voice AI platforms, RAG systems, and LLM orchestration frameworks is highly desirable. You bring exceptional communication skills, customer empathy, and the drive to build AI solutions that transform enterprise operations globally. Requirements What You'll DoCustomer Engagement & Multimodal Agent Development Work directly at customer sites—from factory floors to executive offices—conducting discovery workshops and technical assessments to identify high-impact AI opportunities Design and architect end-to-end multimodal agent systems (voice + video + text) that leverage AION's distributed GPU infrastructure and managed services Build production-grade voice AI systems using STT, TTS APIs, and LLMs deployed on AION's platform Develop vision-enabled agents processing real-time video streams using computer vision pipelines on AION's infrastructure Implement sophisticated multi-agent orchestration with(or similar) frameworks like LangChain or LlamaIndex—enabling tool use, memory management, and autonomous task completion Rapidly prototype POCs in 2-4 weeks, coding alongside client teams to validate concepts and iterate based on feedback Optimize for sub-500ms latency, natural conversation flow, turn detection, and interruption handling in real-time systems Integrate agents directly into customer codebases via REST/GraphQL/WebSocket APIs and custom SDKs (Python, TypeScript) Act as trusted technical advisor to customers, shaping AI strategy and guiding roadmap decisions from concept to production Data Strategy & MLOps Infrastructure Design data architectures with efficient processing pipelines and ingestion workflows for training and inference on AION's platform Implement RAG systems with vector databases—optimizing embedding strategies, chunk sizes, and retrieval methods Prepare and validate datasets for fine-tuning, evaluation, and synthetic data generation Work with other MLEs, MLOps, SREs to carry out model deployment and productionization Observability, Evaluation & Production Operations Implement LLM and agents observability and monitoring—tracking token usage, latency, costs, and quality metrics across deployments on AION's infrastructure Instrument applications to trace LLM calls, retrieval operations, agent actions, and data flows Build evaluation frameworks with offline benchmarks (accuracy, relevance, safety metrics) and online monitoring (user feedback, drift detection) Technical Skills & Experience If you are meeting some of these requirements and feel comfortable catching up on others, we definitely recommend you