Senior Software Engineer, AI Infrastructure - LVM Inference & Evaluation

Build a safer world with us, one incident at a time.

Ambient.ai is the category creator and leader in Agentic Physical Security. Powered by Ambient Pulsar, the first reasoning Vision‑Language Model purpose‑built for physical security, our platform seamlessly integrates with existing security cameras and physical access control systems to unify monitoring, access control, threat assessment, response, and investigations through an always‑on reasoning layer that augments security operators with superhuman capabilities. The results: 95% fewer false alarms, investigations 20x faster, and 10x faster response.

The momentum speaks for itself: we doubled new ARR in FY26, we process 200M+ video hours per day, and have delivered results for world‑class customers including Cisco, Service Now, Sentinel One, Tik Tok, Bayer, and MoMA. That kind of momentum creates an environment where great people thrive, and it shows: we recently ranked #71 out of 500 on the Forbes best startup employers list.

Founded in 2017 and backed by Andreessen Horowitz, Y Combinator, and Allegion Ventures, Ambient.ai is on a fast‑paced journey to fulfill our mission: prevent every security incident possible.

Ready to learn more? Connect with us on Linked In and You Tube

About the role:

Reporting to Raghu Nallamothu, you will design, build, and optimize the AI infrastructure that powers Ambient.ai’s real‑time intelligence platform.

In this role, you will work on the systems required to run state‑of‑the‑art deep learning models across many terabytes of video data in real time. You will help build and scale infrastructure for inference, evaluation, and continuous model improvement across computer vision models, large language models, large vision models, and multimodal AI systems.

This role is ideal for someone with a strong blend of infrastructure engineering, production ML systems, LLM/LVM inference, evaluation harnesses, and inference optimization experience. You will partner closely with research scientists and product engineering teams to bring the latest AI advancements into production for our customers.

What you’ll do:

• Design, build, and maintain cutting‑edge AI infrastructure for real‑time computer vision, LLM, LVM, and multimodal inference workloads.
• Build scalable systems for running state‑of‑the‑art models across large volumes of video and sensor data.
• Optimize inference performance across latency, throughput, GPU utilization, reliability, and cost.
• Develop robust evaluation harnesses and benchmarking systems to measure model quality, system performance, regressions, and production readiness.
• Build infrastructure for continuous model evaluation, experimentation, and deployment.
• Partner with research scientists to product ionize the latest advances in computer vision, LLMs, LVMs, RAG, and multimodal AI.
• Improve model‑serving architecture, including batching, caching, routing, quantization, model parallelism, and hardware utilization.
• Develop data engines and feedback loops for collecting training data, evaluating model behavior, and continuously improving AI performance.
• Create reliable observability, monitoring, and debugging tools for production AI systems.
• Help define best practices for deploying, evaluating, and operating AI systems in real‑world enterprise environments.

What you’ll bring:

• 4+ years of industry experience building infrastructure, distributed systems, machine learning platforms, or production AI systems.
• BS/MS in Computer Science or a related technical field, or equivalent practical experience.
• Strong programming background, especially in Python, with solid software engineering fundamentals.
• Experience designing and building scalable machine learning infrastructure for training, inference, evaluation, and deployment.
• Hands‑on experience running deep learning models in production, ideally including LLMs, LVMs, vision‑language models, or multimodal models.
• Strong understanding of inference optimization techniques, including batching, caching, quantization, parallelism, memory optimization, GPU utilization, and latency reduction.
• Experience with model‑serving frameworks or systems such as vLLM, Triton Inference Server or similar…