Physical AI platform for autonomous city infrastructure monitoring and self-healing operations
EchoTwin AI operates a physical-AI stack built on Python, TensorFlow, PyTorch, and vision-language models, running inference on edge hardware (NVIDIA Jetson) while orchestrating scale through Airflow, dbt, and cloud data warehouses. The engineering-heavy org (7 of 9 roles) with senior IC dominance reflects deep model work; active projects span image captioning, visual QA, and multimodal edge inference—suggesting the company is past foundational research and focused on operational deployment. Pain points around asset-failure prediction and end-to-end program delivery indicate they're solving for both accuracy and municipal workflow integration.
EchoTwin AI builds a cognitive-city platform that turns municipal vehicle fleets into distributed sensing networks. The system ingests real-time visual data, detects infrastructure hazards and asset defects using AI vision, predicts maintenance failures, and generates work orders routed through compliant operational workflows. A verification loop closes each cycle by re-inspecting locations to confirm repairs and measure outcomes. The platform targets mid-sized and large municipal agencies seeking to shift from reactive, manual inspections to data-driven, auditable service delivery. Founded in 2024, the company is based in Boca Raton, Florida, and operates with minimal recent hiring momentum.
Python, TensorFlow, PyTorch, Hugging Face transformers, OpenCV, NVIDIA Jetson for edge inference, AWS/GCP/Azure for cloud, Airflow + dbt for data pipelines, BigQuery/Redshift/Snowflake for warehousing, and FastAPI/Django for backend services.
Boca Raton, Florida. The company hires engineers and data roles in the United States and Serbia.
Active projects include asset-failure prediction, image captioning and visual question answering, multimodal AI at the edge, transit performance assessment, IoT health monitoring, and a CityView web portal for municipal operators.
Other companies in the same industry, closest in size