Principal AI Software Engineer

Responsibilities

• Design and build AI-powered tools that improve productivity across engineering and non-engineering teams (e.g., developer workflows, customer support tooling, tax development, operations, compliance, and analytics).
• Develop internal AI platforms, shared services, and reusable components that enable teams to safely and effectively build AI-powered applications.
• Build and maintain retrieval-augmented generation (RAG) systems grounded in internal documentation, codebases, tax knowledge, and proprietary data.
• Partner with DevOps, platform, and security teams to ensure AI systems are scalable, observable, secure, and compliant.
• Earn and embed AI capabilities into existing Taxwell products to improve accuracy, usability, personalization, and customer outcomes.
• Architect and build new customer-facing products where AI is a core differentiator, not an add-on.
• Work closely with product, design, and tax domain experts to translate real user problems into effective AI powered solutions.
• Ensure AI features meet high standards for reliability, explainability, and trust—especially in a regulated domain like tax.
• Design and implement agent-based systems using frameworks such as LangChain, LangGraph, ADK, or similar.
• Evaluate and integrate commercial and open-source AI technologies, including frontier models and open-weight LLMs.
• Establish best practices for prompt engineering, evaluation, monitoring, and continuous improvement of AI systems.
• Influence AI architecture and technical direction across the organization through design review and hands-on collaboration.
• Measure system effectiveness using telemetry, automated metrics, and qualitative feedback from internal and external users.

Requirements

• 10+ years of software engineering experience, including 3+ years' operating at a Principal- or Staff-level role.
• Proven experience designing and deploying production AI systems used by real users.
• Strong hands-on experience with LLM-enabled systems, including prompt engineering, RAG, and agent orchestration.
• Proficiency in Python and/or TypeScript.
• Experience integrating AI systems with real-world applications, workflows, and data sources.
• Familiarity with vector databases and embedding techniques (e.g., Pinecone, FAISS, Chroma).
• Experience deploying and operating systems on AWS, GCP, or Azure.
• Strong architectural judgment and the ability to balance experimentation with production rigor.

Qualifications

• Experience building AI platforms or internal copilots used across multiple teams or departments.
• Experience developing customer-facing AI products at scale.
• Background in regulated industries such as fintech, tax, or healthcare.
• Experience fine-tuning or customizing open-weight models (e.g.,GPT-OSS-120B, Mistral 3, GLM-4.7).