Senior Forward Deployed ML Engineer, Agents

About the role

Aion is an enterprise AI platform that builds, fine-tunes and deploys AI at scale, and its teams deploy alongside customers to turn production-ready AI into business outcomes. This role suits a hands-on AI engineer with 3-5+ years of experience who works in small teams to deliver high-stakes customer projects, embedding at client sites to design, build and deploy multimodal agent solutions.

Key responsibilities

• Work at customer sites to run discovery workshops and technical assessments, then design and build multimodal agent systems (voice, video and text) on Aion's distributed GPU infrastructure
• Build production-grade voice AI using STT/TTS APIs and LLMs, and vision-enabled agents with computer vision pipelines
• Implement multi-agent orchestration with frameworks such as LangChain or LlamaIndex, and prototype POCs in 2-4 weeks
• Optimize for sub-500ms latency, natural conversation flow, turn detection and interruption handling
• Integrate agents into customer codebases via REST/GraphQL/WebSocket APIs and custom SDKs (Python, TypeScript)
• Design data architectures, implement RAG systems with vector databases, and prepare datasets for fine-tuning and evaluation
• Deploy models with MLEs, MLOps and SREs, and set up observability and evaluation frameworks tracking token usage, latency, costs and quality

Required profile

• 6-8+ years of hands-on experience building production AI/ML systems, with 3-4+ years deploying LLM applications
• Experience building voice agents, vision systems or conversational AI, and with LangChain, LlamaIndex or similar frameworks
• Production-grade Python and JavaScript/TypeScript, RAG with vector databases, and Docker, Kubernetes, CI/CD and infrastructure-as-code
• Experience with AWS, Azure or GCP; customer-facing experience in Solutions Architecture, Technical Account Management or Pre-Sales is highly desirable

Required skills

• Preferred: voice AI platforms, WebRTC and real-time streaming, model fine-tuning (LoRA/QLoRA, supervised fine-tuning, RLHF), inference optimization (vLLM, TensorRT-LLM, Triton, quantization), and LLM observability tools

What we offer

• Equity reflecting your contributions
• Competitive compensation and flexible work options
• Wellness benefits