Remote AI Application Engineer

Required Skills & Experience

• 3+ years of experience in software engineering, application engineering, platform engineering, or software architecture.
• 2+ years of experience developing and deploying machine learning, generative AI, or LLM-enabled applications.
• Strong proficiency in Python, including experience developing APIs and backend services.
• Experience building production-grade integrations using REST APIs, service-based architectures, or event-driven workflows.
• Hands-on experience integrating large language models into applications.
• Experience developing AI agents, multi-step workflows, or agentic applications.
• Experience with tool and function-calling architectures.
• Experience with Model Context Protocol (MCP), including building or integrating MCP servers, tools, or clients.
• Experience with AI agent and orchestration frameworks such as: ○ Pydantic AI ○ LangGraph ○ LangChain ○ Semantic Kernel ○ LlamaIndex ○ Comparable agent-development frameworks
• Experience working with LLM ecosystems such as: ○ Anthropic Claude ○ OpenAI ○ Azure OpenAI ○ Google Gemini ○ Comparable commercial or open-source models
• 2+ years of experience working with Google Cloud Platform.
• Experience deploying containerized applications using Docker.

Nice to Have Skills & Experience

• Integration testing and stress testing.
• Proficiency with FastAPI or a similar modern Python API framework.
• Drug discovery, biology or life science background or work experience.

Job Description

One of our leading pharmaceutical clients is seeking an AI Application Engineer to design and build AI-enabled applications supporting early-stage drug discovery and scientific research. This role will focus on integrating an AI reasoning environment with scientific data and visualization platforms to create a more unified experience for researchers. The engineer will build the application and integration layer that allows scientists to use large language models and intelligent agents while maintaining appropriate controls around data provenance, metadata, access permissions, dataset versioning, and reproducibility. This is not a traditional AI/ML model-development position. The ideal candidate will bring a strong software or platform engineering background and experience building production-grade AI applications, APIs, integrations, and agentic workflows. The engineer must be comfortable working with scientists and domain experts to translate complex research needs into reliable, scalable technical solutions.

Key Responsibilities

• Design, develop, and deploy production-grade AI applications and agent-based systems supporting scientific research and early drug discovery.
• Integrate AI reasoning capabilities with scientific-data, research, and visualization platforms.
• Build APIs, services, and integration layers that allow AI applications to securely access governed scientific data.
• Develop applications using large language models, agent frameworks, tool-calling architectures, and orchestration patterns.
• Build and maintain Model Context Protocol (MCP) servers and tools that make scientific datasets and platform capabilities accessible to AI agents.
• Design workflows that preserve critical data-governance information, including: ○ Data provenance and source traceability ○ Dataset metadata ○ User permissions and access controls ○ Dataset and API version visibility ○ Identification of the curated or approved data source
• Develop integration approaches for environments where traditional APIs may not be available, including automated and version-controlled data feeds.
• Support semantic discovery across scientific datasets, helping researchers identify: ○ What data is available ○ What data they are permitted to access ○ Which datasets are relevant to a scientific question ○ Which sources and dataset versions are being used
• Build observability and reliability into AI applications through appropriate logging, tracing, telemetry, error handling, retry mechanisms, and failure recovery.
• Capture agent activity, orchestration steps, tool calls, source data, prompts, model responses, and dataset versions to support reproducibility and explainability.
• Help investigate situations where an AI-generated scientific interpretation differs from the interpretation of a human expert.
• Develop testing and validation approaches for AI workflows, platform integrations, data pipelines, and production failure scenarios.
• Containerize and deploy services within Google Cloud environments.
• Partner with scientists, data engineers, software engineers, AI/ML engineers, and other domain experts to translate scientific questions into technical workflows.
• Communicate technical concepts clearly to stakeholders who may not have deep software engineering or AI backgrounds.
• Contribute to architectural decisions regarding direct platform-to-platform integration versus the introduction of an intermediary application or orchestration layer.