The Mandate

You are responsible for building the “command center” that understands the physical world. What cameras and sensors transmit back are not just cold pixels but events from the physical world. You need to fuse these event streams — spanning time, space, and devices — into a massive, inferable “Reality Graph,” seamlessly integrating LLMs, RAG, and graph databases.

Core Tech Stack

Primary Languages: 

Proficient in Python / Golang.

Graph and Vector Engines: 

Must have hands-on experience building Knowledge Graphs using graph databases (Neo4j, NebulaGraph); proficient in vector databases and hybrid search.

AI Architecture: 

Deep understanding of LLM principles, skilled in building RAG systems and multi-agent collaboration workflows.

Multimodal: 

Familiar with both Computer Vision and NLP, knowing how to transform unstructured data into structured graph nodes.

Bonus (Strongly Preferred)

Experience building a knowledge graph with tens of billions of nodes from scratch.

Deep understanding of time-series databases and event stream processing (Kafka, Flink).

Company Website: https://kamivision.com/en-us