Job Title: Software Engineer, AI/ML
Location: San Francisco, CA – In-person 3 days per week
Team: Engineering
Reports To: Head of Engineering
Company: Marble Tax

About Us

Marble Tax is building an AI-powered tax assistant and preparation platform that transforms how accounting firms and their clients handle tax season. We’re targeting the pain points of mid to large-sized accounting firms with intelligent tools designed to streamline complex tax workflows. We’re early stage, highly collaborative, and deeply focused on solving real, high-impact problems for our customers.

About the Role

We’re looking for a multi-faceted AI/ML-focused engineer who thrives in early-stage environments and is eager to work across the stack — from building LLM-powered agentic systems to designing traditional ML pipelines that enhance performance and accuracy. You’ll own key technical decisions, build production-ready AI/ML features, and work closely with the rest of the engineering and product team to define the architecture, product experience, and team culture. This is a foundational hire with a clear path toward leading a team of AI engineers, researchers, and data scientists. Early on, you’ll be hands-on — prototyping, experimenting, and shipping — but you’ll also influence the long term platform, team, and technical strategy.

What You’ll Do

• Design and build core AI-driven systems, including LLM-based retrieval-augmented generation (RAG) pipelines, agent orchestration, and prompt engineering frameworks
  

• Lead development of classifiers, validators, re-rankers, and other traditional ML components to improve accuracy, reduce errors, and enhance decision-making in our assistant workflows
  

• Build robust document parsing and information extraction pipelines capable of handling complex, unstructured tax-related documents and updating workflows retroactively as new data arrives
  

• Collaborate tightly with product, engineering and leadership to quickly test, validate, and improve AI-driven features based on user feedback
  

• Mentor peers, contribute to hiring decisions, and help establish best practices for AI/ML development, evaluation, and deployment
  

• Balance research with production, delivering reliable, secure, and maintainable systems while staying informed on emerging AI/ML developments

What We’re Looking For

• Proven experience building and deploying AI/ML systems in production, ideally in a product-focused or startup environment
  

• Strong technical foundation in applied ML and software engineering
  

• Deep understanding of LLMs, RAG systems, and agent frameworks, with experience optimizing for latency, accuracy, and cost
  

• Experience with traditional ML techniques, including text classification, ranking, entity extraction, and anomaly detection
  

• Ability to work with unstructured data, build scalable parsing pipelines, and integrate multiple data sources into coherent workflows
  

• Comfort with ambiguity and wearing multiple hats: engineer, researcher, product collaborator, and mentor
  

• Excellent communication skills and a collaborative approach to problem-solving
  

• Comfortable being in person 3 days a week in a highly interactive environment
  

Bonus Points

• Experience with AI evaluation frameworks, fine-tuning LLMs, or building custom embeddings
  

• Background in accounting, tax software, or enterprise SaaS
  

• Prior experience growing and leading AI/ML teams

Why Join Us?

• Shape the technical and product direction of an AI-first platform from the ground up
  

• Work at the forefront of applied AI, solving real and high-value problems
  

• Build from 0 -> 1 alongside a team that values autonomy, pragmatism, and kindness
  

• Competitive compensation, equity, and benefits

Interested?
We’d love to hear from you. Whether you’re currently leading teams or an engineer ready to step up, let’s talk.