Artificial Intelligence Engineer

Role Description
The AI Engineer is responsible for the engineering execution and delivery of the machine learning systems behind Simplifisign's verification platform. This role owns the models and services that power identity and entity verification across KYC, KYB, and KYV workflows, and is accountable for their accuracy, latency, and auditability in production.
The AI Engineer partners closely with product and compliance stakeholders to translate verification requirements into technical solutions, and works alongside the platform engineering team to integrate model inference directly into live customer workflows. This role plays a key part in turning verification policy into executable pipelines and ensuring those pipelines hold up under regulatory scrutiny.
Establish the evaluation and monitoring layer for every deployed model, including offline evaluation harnesses, drift detection, shadow deployments, and production metric alignment
Qualifications
University degree/diploma in Computer Science, Engineering, or equivalent practical experience
3+ years of software, data, or ML engineering experience, with at least 3 years shipping models that other production systems depend on
Strong Python and PyTorch, plus the ability to write and own service code rather than handing off a model artifact
Hands on experience with transformer based document understanding, such as LayoutLM, Donut, or comparable architectures, and with classical OCR pipelines
Working knowledge of Go, gRPC service design, and Postgres schema and query performance
Demonstrated experience with imbalanced classification, calibration, and threshold selection under real business cost constraints
Familiarity with model deployment, versioning, and monitoring in a containerized environment
Preferred qualifications:
Graph modelling and entity resolution experience, ideally with Neo4j or a comparable graph database
Background in fintech, RegTech, identity verification, or another regulated domain
Exposure to AML, sanctions screening, or adverse media matching
Experience with retrieval augmented generation or LLM based extraction where deterministic parsing is insufficient

Work arrangement
No