Scientific Software Developer

Enthought, Inc.; Austin (TX); Hybrid; USD 120,000 - 180,000; Full time

Benefits offered by this job

Meaningful impact

Agentic AI focus

World-class colleagues

Continuous learning

Global offices: Austin, Cambridge, and

Flexible hybrid work

Competitive compensation and benefits

Job summary

Enthought, Inc. in Austin, TX, is hiring a Scientific Software Developer to join client-facing project teams that accelerate R&D decisions and bring models to production.

The work spans data exploration through model development to production software built on Enthought’s proprietary technology and the open-source scientific Python ecosystem. Agentic AI coding tools are the default, so you will direct agents, check outputs, and own the result while sitting with client researchers and leaders to

Responsibilities

• Design end-to-end scientific solutions, applying deep knowledge of statistics, optimization, and machine learning to translate client R&D problems into architectures spanning data, models, and user-facing applications.
• Work agent-first, using AI coding assistants and autonomous coding agents as your default way to explore data, prototype models, build software, write tests, and produce documentation.
• Direct AI agents like a technical lead. Decompose scientific and engineering problems into well-scoped tasks, supply the domain context and scientific constraints, write clear specifications, and iterate toward correct, defensible results.
• Evaluate what the agents produce. Catch errors, hallucinations, security issues, and subtle scientific or statistical inaccuracies before they reach a client.
• Build data-driven models, AI-based decision-support tools, and automated analysis pipelines, then assemble them into web-based solutions involving GUIs, 2D and 3D graphics, and cloud computing.
• Design AI-enabled and agentic capabilities into client solutions where they add value: LLM-powered literature and data extraction, tool and function calling, retrieval over scientific data, and evaluation harnesses, held to the same rigor as any other component.
• Apply judgment about where generative approaches fit and where they do not, particularly with the small, sparse, and noisy datasets typical of scientific R&D, balancing speed against reliability and reproducibility.
• Participate in the team’s software practices: release planning, architecture reviews, code review, sprint planning, retrospectives, and standups.
• Interface with clients to define, demonstrate, and refine solutions, and handle their data and intellectual property responsibly, using AI tools only in approved, secure configurations.

Job description

The Opportunity

We are hiring a Scientific Software Developer to work on client-facing project teams that help industrial R&D organizations make better, faster decisions and get products to market sooner. The work runs from data exploration through model development to production software, built on Enthought’s proprietary technology and the open-source scientific Python ecosystem. Agentic AI coding tools are the default way the work gets done here, so you will spend your days directing agents, checking what comes back, and owning the result. You will sit with client researchers and business leaders often enough that explaining a method clearly matters as much as building it.

Key Responsibilities

• Design end-to-end scientific solutions, applying deep knowledge of statistics, optimization, and machine learning to translate client R&D problems into architectures spanning data, models, and user-facing applications.
• Work agent-first, using AI coding assistants and autonomous coding agents as your default way to explore data, prototype models, build software, write tests, and produce documentation.
• Direct AI agents like a technical lead. Decompose scientific and engineering problems into well-scoped tasks, supply the domain context and scientific constraints, write clear specifications, and iterate toward correct, defensible results.
• Evaluate what the agents produce. Catch errors, hallucinations, security issues, and subtle scientific or statistical inaccuracies before they reach a client.
• Build data-driven models, AI-based decision-support tools, and automated analysis pipelines, then assemble them into web-based solutions involving GUIs, 2D and 3D graphics, and cloud computing.
• Design AI-enabled and agentic capabilities into client solutions where they add value: LLM-powered literature and data extraction, tool and function calling, retrieval over scientific data, and evaluation harnesses, held to the same rigor as any other component.
• Apply judgment about where generative approaches fit and where they do not, particularly with the small, sparse, and noisy datasets typical of scientific R&D, balancing speed against reliability and reproducibility.
• Participate in the team’s software practices: release planning, architecture reviews, code review, sprint planning, retrospectives, and standups.
• Interface with clients to define, demonstrate, and refine solutions, and handle their data and intellectual property responsibly, using AI tools only in approved, secure configurations.

What We Offer

• Meaningful impact. Your work directly advances scientific breakthroughs, from accelerating drug discovery to developing sustainable materials.
• A front-row seat to agentic AI. We are an automation-first company where every role directs AI tools, and you will help define how the industry puts them to work.
• World-class colleagues. You will work alongside some of the smartest, kindest scientists and engineers around.
• Continuous learning. Enthought’s training programs in Python, machine learning, and scientific computing are open to you.
• A global, collaborative culture across our Austin, Cambridge, and Tokyo offices.
• Flexible hybrid work based out of our Austin office.
• Competitive compensation and benefits.