Principal Applied AI Engineer — Agentic Intelligence

Principal Applied AI Engineer — Agentic Intelligence

SenseMesh is building a real-time AI operating system for the physical world.

We connect cameras drones and other sensors into a shared reality graph — a live world model that helps people and machines understand what is happening reason across time and space and decide what to do next. Most systems in this category still stop at recording detection or alerting. We are building something much more capable: a system that helps operators investigate understand decide and increasingly automate what happens next.

We already have strong foundations in perception messaging infrastructure and an emerging intelligence layer on top. Now we are looking for a senior hands-on engineer to take ownership of that layer and help turn it into one of the defining capabilities of the company: natural language reasoning agentic search and evidence-grounded intelligence over a live multi-sensor world model.

‍You will work closely with the founding team — especially the CTO and CPO — and the broader engineering team to define how users ask questions of the physical world how the system answers with evidence and judgment and how perception turns into deeper understanding and action.

What you’ll work on

You’ll help build systems that let users ask questions like:

• What happened at the front door between 2 and 4 PM?
• Summarize everything unusual that happened overnight.
• Is anyone home right now?

Your work will include

• natural language search and reasoning over a multi-camera multi-sensor world model
• retrieval systems that combine structured filters vector similarity graph traversal ranking and summarization
• routing and fusing sub-queries across heterogeneous sources
• evolving retrieval toward direct visual embeddings such as CLIP and SigLIP
• temporal reasoning over durations sequences and event order
• evaluation methods for open-ended investigative queries where ground truth is often incomplete or ambiguou

What we’re looking for

Strong candidates will usually have built production systems in several of these areas:

• AI search question answering or assistant systems
• retrieval ranking and multi-step query orchestration
• agent or tool-use workflows
• graph-backed or structured-data reasoning
• multimodal or video understanding
• evals observability and iterative quality improvement

Above all we are looking for someone with exceptional coding ability strong engineering judgment and strong product instincts — someone who can turn ambitious AI ideas into systems people trust in real-world conditions.

Why this role matters

We are not hiring someone to add AI to the product.

We are hiring someone to help define the intelligence layer that makes the product extraordinary.

If you want a place to do some of the best and most consequential work of your career this is that kind of role.

Let’s build something great together.

[email protected]

What We Do

SenseMesh builds AI systems that understand and coordinate the physical world in real time.Our platform connects cameras drones microphones IoT devices and other sensors into a unified AI operating layer. Instead of forcing operators to stare at dashboards or search through hours of footage SenseMesh turns raw sensor data into structured searchable human-narratable intelligence — what we call the Reality Graph.Customers can ask natural-language questions about what happened define missions for what the system should watch for and trigger customer-approved workflows when important conditions occur. The system is designed to move beyond simple object detection into temporal reasoning: understanding events sequences persistence anomalies and context over time.We are starting with high-stakes physical environments — schools hospitals public safety infrastructure venues and other places where real-time understanding matters.

Why Work With Us

Most video AI companies are still building better tags dashboards and alerts. SenseMesh is building something more fundamental: an agentic intelligence layer for the real world that understands what is happening and helps customers take action in real time.