TensorFlow Model & NiFi Integration
Publicada el 2026-07-24
Descripción de la oferta
This project has two parallel tracks. First, a fresh machine-learning model must be built in TensorFlow. I will supply sample data and the business objective as soon as the engagement starts; your task is to design the architecture, train the model, and hand over reproducible training code together with saved weights and concise documentation. Second, the model has to fit smoothly into an existing Java + Vert.x micro-service landscape. The runtime environment already speaks to “any” message queue (RabbitMQ is in place, yet the design should stay broker-agnostic), so the model-serving endpoint has to publish and consume messages accordingly. Apache NiFi + Apache Camel will orchestrate inter-service communication and system-to-system integration, and you need to have used both so you can move fastest with different iterations of the project. Deliverables • TensorFlow training notebook or scripts, trained model artefacts, and a brief README. • A small Vert.x service exposing the model over REST and the MQ bus. • NiFi flow or Camel routes that integrate the service with upstream and downstream systems, packaged for straightforward import. • A short hand-off note explaining deployment and configuration. When you reply, focus on the experience you have with these exact tools—TensorFlow, NiFi or Camel, Java Vert.x, and message-queue based architectures. Code quality, clarity, and reproducibility will be the acceptance criteria for sign-off.
Skills
Fuente original: freelancer