Applied AI Engineer (LLM)
New
S
SportyApplied AI
Location: EMEA, Our core working hours are 10am-3pm in your local time zoneFull-TimeMiddle
Salary not disclosed
Apply NowOpens the employer's application page
Job Details
- Languages
- Ability to discuss technical designs and document results in English
- Experience
- 3+ years of experience in software engineering, machine learning engineering or applied AI.
- Required Skills
- PythonMachine LearningRESTful APIs
Requirements
- Have 3+ years of experience in software engineering, machine learning engineering or applied AI.
- Have delivered an LLM application or service used by real users, with clear evidence of your contribution and results.
- Demonstrate strong Python skills.
- Have experience building APIs, processing data and debugging software.
- Have hands-on experience in at least one of: LLM application development, model serving, model evaluation and fine-tuning, or coding-agent development.
- Be able to design repeatable evaluations, investigate failures and explain trade-offs between quality, response time and cost.
- Be able to discuss technical designs and document results in English when collaborating with international teams.
- Experience operating open-weight models with frameworks such as vLLM or SGLang, including load testing and GPU memory optimisation, is an advantage.
- Experience fine-tuning models with domain-specific data and validating improvements on test data kept separate from training is an advantage.
- Experience building document-grounded retrieval systems or coding agents integrated with repositories and automated tests is an advantage.
Responsibilities
- Build internal AI services using open-weight language models for document processing, customer support and employee chat.
- Connect models to company systems and approved data through APIs, implementing output validation, access controls, failure handling and fallback to approved providers.
- Work with Infrastructure and DevOps teams to deploy models on company-controlled infrastructure.
- Improve response times, concurrent request capacity and serving costs using representative workloads.
- Compare internal models with existing providers through repeatable quality, latency and cost evaluations.
- Use evaluation results to guide model selection, prompts, retrieval and fine-tuning, and verify quality during staged rollouts.
- Integrate internal models with coding assistants and agents that retrieve code, propose changes and run tests within approved permissions.
- Help teams adopt coding tools through practical examples and training while preserving code-review and release requirements.
- Track usage by business and service, compare operating costs with estimated external-provider costs, and measure workflow time saved, defect rates, review and rework.
View Full Description & ApplyYou'll be redirected to the employer's site