ML / AI Engineer

About the role We need an ML engineer to take ownership of a cloud-based AI pipeline currently running in production with no backup. The previous engineer is departing. The pipeline covers OCR, document extraction, voice ID, translation, and transcription - all deployed as REST APIs. What you'll work on - Cloud OCR pipeline - classifying uploaded images (invoices, logbooks, weighing scales, licence plates), extracting structured data using Mistral OCR, and returning results to mobile and web clients - Transaction verification - matching multiple image types into coherent transaction records - Voice ID / speaker verification - enrolment and probing API using NVIDIA TitaNet - Translation and transcription pipelines - server-side audio transcription and multilingual translation (currently Gemini Flash) - Image classification - routing uploaded images to the correct processing pipeline - API maintenance and deployment - all of the above run as live REST APIs on AWS EC2 Profile A hands-on ML engineer who has built and shipped deployed AI pipelines. The work is applied: wiring models into APIs, handling messy real-world inputs (blurry photos, noisy audio, handwritten text), and keeping things running in production. - Experience deploying ML models as REST APIs - Computer vision / OCR experience - Speaker verification or audio ML is a strong plus - Python (primary); ONNX familiarity a bonus - Comfortable inheriting and owning an existing codebase - Reliable, async-friendly, proactive communicator Stack Python, FastAPI, Mistral OCR, NVIDIA TitaNet, Gemini Flash, AWS EC2.