AI Product & Research Engineer, Capital.com

• We’re spinning up a brand-new AI department and need a first-on-the-ground engineer who can turn wild LLM ideas into working demos overnight, teach everyone around you why automation matters, and have fun doing it
• You can ship Python or TypeScript prototypes
• Master’s degree in AI and/or Computer Science
• Hands-on experience integrating LLM APIs (e.g. OpenAI, Hugging Face Inference)
• Practical experience fine-tuning LLMs via OpenAI, HuggingFace or similar APIs
• Strong proficiency in Python
• Deep expertise in prompt engineering and tooling like LangChain or LlamaIndex
• Proficiency with vector databases (Pinecone, FAISS, Weaviate) and document embedding pipelines
• Proven rapid-prototyping skills using Streamlit or equivalent frameworks for UI demos
• Familiarity with containerization (Docker) and at least one orchestration/deployment platform
• Excellent communication and ability to frame AI solutions in business terms

Desirable

• Prompt Engineering & Orchestration: Craft, iterate, and optimize prompts for APIs (OpenAI, Cohere, Anthropic). Build multi-step “chains” using LangChain, LlamaIndex, or custom controllers. Develop and maintain AI microservices using Docker, Kubernetes, and FastAPI, ensuring smooth model serving and error handling
• Vector Search & Retrieval: Implement retrieval-augmented workflows: ingest documents, index embeddings (Pinecone, FAISS, Weaviate), and build similarity search features
• Rapid Prototyping: Create interactive AI demos and proofs-of-concept with Streamlit, Gradio, or Next.js for stakeholder feedback
• MLOps & Deployment: Implement CI/CD pipelines (e.g., GitLab Actions, Apache Airflow), experiment tracking (MLflow), and model monitoring for reliable production workflows
• Cross-Functional Collaboration: Participate in code reviews, architectural discussions, and sprint planning to deliver features end-to-end