Senior Applied AI Engineer at Harvey

New York, New York, United States

Harvey Logo
$201,000 – $264,000Compensation
Senior (5 to 8 years)Experience Level
Full TimeJob Type
UnknownVisa
Legal Tech, Professional ServicesIndustries

Requirements

  • 7+ years of experience (post-BS/MS) in an engineering role
  • Experience with shipping a scaled and impactful product powered by machine learning: how to use offline datasets, online experiments, and recent research to build simple and high performance systems. Prior experience with LLMs and retrieval pipelines is not required
  • Track record of shipping reliable products and a strong attention to detail
  • Experience building backend platforms that can support multiple product lines
  • Grit - experience working at early-stage startups is a plus

Responsibilities

  • Conduct data collection, experimentation, and analysis to drive algorithmic development for RAG and multi-step AI pipelines
  • Zero-to-one product development: rapidly prototype, evaluate, integrate, and test new product features in close partnership with our legal team
  • Develop new AI native workflows: implement streaming, long-running tasks, procedural UX, etc. for new AI tasks, finding the balance between state-of-the-art and pragmatism
  • Partner with our legal team to design and implement a method for evaluating the correctness of document citations. Use this to build a large dataset of ground-truth citations and then improve our core citation algorithm
  • Make targeted improvements to our RAG pipelines to improve answer quality for user questions over corpuses of complex data, like massive banks of spreadsheets or Japan’s tax code
  • Design and build systems that leverage state-of-the-art LLMs from multiple model providers, including custom models
  • Work across the stack and with our legal team to create seamless multi-step AI workflows for complex legal tasks, like corporate merger due diligence

Skills

RAG
AI Pipelines
Machine Learning
Data Collection
Experimentation
Data Analysis
Prototyping
Full Stack Engineering
LLM Integration

Harvey

Custom AI solutions for law firms

About Harvey

Harvey.ai builds custom Large Language Models (LLMs) specifically designed for top law firms to help them tackle complex legal challenges. These AI models are tailored to various legal practice areas and jurisdictions, allowing firms to enhance their efficiency and accuracy in legal work. Harvey.ai's technology includes an AI chatbot developed in collaboration with Allen & Overy, which demonstrates how their solutions can streamline operations and reduce manual workloads. The company operates on a business model that combines customization fees for developing these models with ongoing subscription fees for support and updates. Unlike many competitors, Harvey.ai focuses exclusively on the legal sector, ensuring that their products meet the unique needs of elite law firms. The goal of Harvey.ai is to transform the legal industry by providing advanced AI tools that improve decision-making and operational efficiency while maintaining high standards of data security.

San Francisco, CaliforniaHeadquarters
2022Year Founded
$200.4MTotal Funding
SERIES_CCompany Stage
Cybersecurity, LegalIndustries
201-500Employees

Risks

Over-reliance on partnerships could lead to strategic vulnerabilities if they dissolve.
LexLegis.ai's global expansion poses a competitive threat in legal research.
Rapid AI innovation and crowded CLM market may challenge Harvey's growth.

Differentiation

Harvey offers custom LLMs tailored for elite law firms' complex legal challenges.
The company provides a unified interface for legal workflows, enhancing efficiency.
Harvey's BigLaw Bench evaluates AI accuracy in legal tasks, setting industry standards.

Upsides

Partnership with Icertis enhances contract analysis and risk assessment capabilities.
Expansion into Japan with Mori Hamada Matsumoto opens new revenue streams.
Revamped GenAI Assistant improves user experience and output quality for legal professionals.

Land your dream remote job 3x faster with AI