The posting
OREDATA is a Digital Transformation & IT Consulting firm with 10+ years of proven expertise and hundreds of successfully implemented projects across the EMEA region. When you join the OREDATA team, you'll be working hand-in-hand with experts focused on tackling digital, operational, analytical & data science challenges with the greatest impact. We foster collaboration with proximity, an agile and autonomous approach and best practices and guiding principles.
We are looking for an Data Scientist (LLM) to join our team within a leading company in the aviation industry. ✈️
Read more: https://medium.com/@oredata-engineering
Apply and be part of our exciting journey!
Responsibilities
- Design, develop, test, and productionize LLM-based and Retrieval-Augmented Generation (RAG) solutions for enterprise use cases.
- Take an active, hands-on role in the development of internal chatbot, conversational AI, knowledge assistant, and agentic AI products, from POC/MVP through production readiness.
- Own and contribute to the technical architecture of enterprise LLM solutions, including model selection, deployment, serving, routing, evaluation, monitoring, and integration with internal AI platforms and applications.
- Deploy, operate, and optimize open-weight and commercial LLMs, with a particular focus on on-premise and private infrastructure. This includes taking a lead role in standing up and configuring on-premise platforms (such as Red Hat OpenShift AI) from scratch when necessary.
- Evaluate and select appropriate models based on use-case requirements, considering quality, latency, throughput, infrastructure requirements, cost, security, licensing, and operational constraints.
- Optimize LLM inference and infrastructure utilization through techniques such as quantization, batching, caching, model serving optimization, GPU resource management, and appropriate workload allocation.
- Act as an advocate for AI infrastructure efficiency (AI FinOps), optimizing compute costs by balancing model performance, hardware allocation (e.g., Multi-Instance GPU), and semantic routing strategies.
- Design and improve model routing and semantic routing mechanisms to ensure requests are handled by the most appropriate model based on use case, complexity, performance, and resource requirements.
- Design and support agentic and tool-calling architectures, ensuring that the appropriate models, tools, and enterprise services are selected and invoked reliably and securely.
- Contribute to the evolution of the organization’s AI Gateway and shared AI platform capabilities, including model access, authorization, quotas, routing, governance, observability, and usage controls.
- Work with structured and unstructured data to prepare, retrieve, enrich, and optimize knowledge sources used by AI applications, including embeddings, vector search, hybrid retrieval, re-ranking, chunking, and context management.
- Establish and improve LLM evaluation and monitoring practices, including benchmark datasets, offline and online evaluation, regression testing, output quality analysis, hallucination monitoring, and performance metrics.
- Collaborate closely with platform, infrastructure, data, software engineering, architecture, product, and business teams to translate business requirements into scalable and operationally feasible AI solutions.
- Provide technical direction and architectural guidance on GPU capacity, AI infrastructure utilization, model serving technologies, and platform evolution, while remaining actively involved in implementation when required.
- Stay current with developments in Generative AI, LLMs, agentic systems, model serving, inference optimization, RAG architectures, and AI infrastructure, and evaluate their practical applicability within the enterprise environment.
Requirements
Must-Haves (Minimum Qualifications)
- Minimum 7 years of professional experience in Artificial Intelligence, Machine Learning, Data Science, Software Engineering, Data Engineering, AI Platform Engineering, or related technical roles.
- Strong hands-on experience with Large Language Models (LLMs) and Generative AI solutions, including experience taking AI systems beyond experimentation and into production environments.
- Proven experience deploying, serving, operating, or optimizing LLMs, preferably in on-premise, private cloud, or enterprise containerized environments.
- Proven experience designing and scaling AI systems for high-traffic, high-concurrency environments, ensuring latency control and graceful degradation under heavy load (e.g., handling traffic spikes).
- Practical understanding of GPU-based LLM inference and the key factors affecting GPU memory utilization, throughput, latency, concurrency, and infrastructure efficiency.
- Hands-on knowledge of LLM inference optimization techniques such as quantization, batching, caching, model selection, and serving optimization.
- Experience working with open-weight models and model ecosystems/frameworks such as Hugging Face, vLLM, NVIDIA inference technologies, TGI, Triton, or comparable technologies.
- Experience with containerized infrastructure and orchestration technologies such as Kubernetes and/or OpenShift.
- Strong practical experience designing, building, and improving RAG-based applications, including embeddings, vector databases/search, document retrieval, chunking strategies, hybrid retrieval, re-ranking, context management, and retrieval quality optimization.
- Experience with model routing, semantic routing, or multi-model architectures, with the ability to determine how different models should be selected and utilized.
- Hands-on experience with agentic AI workflows, tool/function calling, orchestration patterns, and integration of LLMs with internal/external tools.
- Strong programming skills, preferably in Python, together with solid software engineering practices including testing, API design, version control, CI/CD, code quality, and maintainable system design.
- Strong analytical thinking and problem-solving capability, with the ability to independently investigate technical problems, evaluate alternatives, make technical decisions, and drive solutions toward production.
Nice-to-Haves (Highly Preferred)
- Specific experience with Red Hat OpenShift AI / Red Hat AI platforms is a strong plus.
- Experience with AI Gateway, API Gateway, model gateway, or shared enterprise AI platform architectures.
- Good understanding of LLMOps/MLOps and model lifecycle management, including model versioning, deployment, monitoring, observability, and production governance.
- Experience with chatbot, conversational AI, knowledge assistant, enterprise search, recommendation, intelligent automation, or similar AI-enabled products.
- Understanding of enterprise considerations around model licensing, open-source/open-weight usage, information security, data privacy, access control, governance, and responsible AI.
- Experience providing technical leadership, architecture guidance, design reviews, or mentoring to other engineers.
Get to know us
If you want to know more about us and what we do, then visit our website: www.oredata.com
Why Oredata?
- Open communication, flexibility and start-up spirit
- Learning & Development opportunities for both personal and professional growth
- Opportunity to get company paid Professional Certificates (Google Cloud Platform, Confluent Kafka, etc)
- Access to Online Training Platforms (Udemy, Pluralsight, A Cloud Guru, Coursera, etc.)
- Dynamic work ecosystem where you can take initiative and responsibility
- Opportunity to work on international projects
- Private Health Insurance
- Birthday Leave Policy
Kişisel verileriniz işe alım sürecinin yürütülebilmesi amacıyla veri sorumlusu sıfatıyla şirketimiz Oredata Yazılım A.Ş. tarafından işlenecektir. Kişisel verilerinizin işlenmesi ve haklarınızla ilgili detaylı bilgiye https://oredata.com/personal-data-protection-policy/ bağlantısı üzerinden ulaşabilirsiniz.
