Biorce is a Healthtech company using AI to accelerate drug development, improve patient outcomes, and expand access to cost-effective treatments. Its team combines clinical research professionals, data scientists, and AI experts to connect advanced technology with real-world clinical needs.
ai engineer for drug development
подготовьтесь к отклику
ai-инструментыЧтобы адаптировать резюме под вакансию или составить сопроводительное письмо, загрузите резюме
описание
задачи
- Design and implement complex, multi-step AI workflows using LangGraph or similar frameworks to manage state, loops, and conditional logic;
- Develop, test, and version-control advanced prompts, Few-shot, Chain-of-Thought, and ReAct;
- Use Vertex AI’s Help me write and prompt optimization tools to refine system instructions;
- Build and maintain Retrieval-Augmented Generation (RAG) pipelines, including document parsing, embedding optimization, and vector database management;
- Establish automated evaluation frameworks using the Vertex AI Gen AI Evaluation Service to measure hallucination rates, groundedness, and relevancy.
требования
- 2/3+ Years of experience in software or ML engineering;
- At least 1 year of hands-on experience building and deploying LLM-based applications in production;
- Experience with orchestration tools such as LangGraph, Stateful Actors, LangChain, CrewAI, or Vertex AI Agent Builder;
- Experience with the GCP ecosystem, including Vertex AI Studio, Cloud Run, BigQuery, and Cloud Storage;
- Advanced RAG techniques, Hybrid search, Parent-Document Retrieval, and Query expansion;
- Deep understanding of conversational state and memory in complex, non-linear AI graphs, cycles, nodes, and edges;
- Ability to apply software engineering practices to AI engineering, including CI/CD, prompt unit testing, and structured logging with Cloud Trace and Cloud Logging;
- Nice to have: experience designing agentic workflows using LangGraph or similar frameworks, prompt engineering and optimization including Few-shot, Chain-of-Thought, and ReAct, building and maintaining RAG pipelines, establishing LLM-as-a-Judge evaluation frameworks, and using Vertex AI Gen AI Evaluation Service to measure hallucination rates, groundedness, and relevancy.
условия
- Comprehensive private health coverage;
- Company events;
- MacBook provided;
- Pet-friendly office.
навыки
Если просят войти через iCloud, отправить коды из SMS, запустить код, что-то установить, перевести деньги или сделать что угодно, связанное с деньгами, не соглашайтесь: это признаки мошенничества.