কাজের বিবরণ

About Orcrist Technologies

Orcrist develops the Orcrist Intelligence Platform (OIP), a Kubernetes-based data intelligence solution offered as SaaS or self-hosted/on-prem (including air-gapped setups). The platform integrates data processing, machine learning/AI, and contemporary web applications to support critical clients in both public and private sectors.

Role Overview

The role involves incubating and validating new machine learning projects from start to finish. Working within the innovation team, you will create ready-to-adopt prototype vertical slices that encompass data processing flows, model deployment, evaluation metrics, and product integration. These prototypes will be clearly documented and passed to delivery teams for long-term production maintenance.

Key Responsibilities

Develop ML prototype vertical slices linking data ingestion, processing, model inference, and visible product features like search, insights, and UX workflows.
Design evaluation frameworks and decision artifacts, including datasets, baselines, metrics on quality, latency, cost, and make actionable go/no-go recommendations.
Prepare prototypes for adoption by containerizing services, defining reproducible deployment methods, and creating runbooks and checklists.
Collaborate with Research and Data Engineering teams on dataset curation, annotation loops, experiment tracking, and iterative safe development.
Ensure prototypes are operationally reliable with instrumentation, monitoring, and compliance principles such as PII handling and data provenance.

Candidate Profile

At least 3 years of experience in ML engineering or MLOps with a proven track record of deploying real-world systems.
Proficient in Python and hands-on experience with PyTorch and Transformers; capable of evolving models from notebooks to reproducible service deployments.
Experience with Kubernetes and container technologies for deploying and troubleshooting clusters, including offline and air-gapped environments.
Strong focus on evaluation and monitoring with the ability to clearly communicate trade-offs and results.
Eligibility to work in Germany is required; preference for EU/NATO citizens due to export control requirements.

Preferred Skills and Experience

Expertise in GPU serving and optimization techniques such as Triton, KServe, ONNX, TensorRT, batching, and quantization.
Knowledge of streaming and pipeline tools like Kafka, Ray, Beam, Flink, or Spark and experience with search, vector, or graph data integrations.
German language skills at B1 level or higher and/or experience working with regulated or public sector datasets and workflows.

Benefits and Work Environment

Engage with a modern ML technology stack operating under real-world constraints including Kubernetes, streaming, and hybrid/on-premise/air-gapped deployments.
Work remotely across Germany with regular workshops in Berlin, 30 days of vacation annually, plus budgets for equipment and professional learning.
Have high impact through your prototypes and handoffs that accelerate multiple delivery teams' progress.

ML Engineer