The Lead Software Engineer will build software solutions that address key business needs across Advita’s global organization using large language models and AI principles. This role will collaborate with functional teams to define business requirements, identify appropriate opportunities for enhanced tools and AI, and design, build, and deploy solutions.
The role is responsible for setting technical direction, writing production code, and supporting solutions after they go live.
Key Responsibilities
• Collaborate with functional teams across any area of the business to understand their requirements, define the problem to be solved, and co-develop the solution.
• Assess where a large language model is the right tool and where conventional software or a data solution serves the business better.
• Design and build retrieval-augmented generation (RAG) solutions, including embedding, vector search, and orchestration, and select the models and services behind them.
• Build evaluation and monitoring that gates each release and tracks quality and cost once solutions are in service.
• Design for auditability, working with quality and regulatory stakeholders so that solutions hold up under inspection.
• Build the integrations and data pipelines that these solutions depend upon, working against enterprise systems of record.
• Own cloud architecture, cost management, security, and access control on Azure.
• Partner with the enterprise applications team so that these solutions reach the people who need them.
Skills Knowledge and Expertise
Education:
• Bachelor's degree in Computer Science, Engineering, a quantitative field, or equivalent practical experience.
Experience:
• 5+ years building and shipping production software.
• Experience delivering production applications built on large language models, including retrieval-augmented generation (RAG), prompt and context design, and evaluation.
• Experience integrating with enterprise systems of record and working with the data they hold.
• Experience building and deploying machine learning models on proprietary data. Preferred.
• Experience in a regulated industry such as medical device or life sciences. Preferred.
• Experience with document-heavy and unstructured data problems, including extraction, classification, and coding. Preferred.
Functional/Technical Knowledge, Skills and Abilities Required:
• Python, Node.js, and TypeScript.
• Hands-on experience with vector databases and embedding models.
• Hands-on experience on Azure, including Azure AI Foundry, along with data services, compute, identity, and networking.
• Judgment about when a large language model is the right solution and when it is not.
• Cost consciousness — weighing model, token, and infrastructure cost as part of the design, and keeping it under control once a solution is running.
• Written and verbal communication skills – ability to present potential solutions, pros and cons to senior stakeholders.
• Collaborative mindset with ability to work directly with stakeholders across the business.
• Able to run several efforts at once without dropping quality.
• Project management: scoping, sequencing, estimating, and reporting progress against commitments.