SoFi Verified

FinTech, Financial Services

AI Engineer

San Francisco, California, United StatesRemoteFull Time$105,600–$198,000 /yrPosted 2 months agoVisa sponsorship available

Is this role right for you?

Upload your resume and get a skill-by-skill breakdown — see exactly where you match, where you're close, and what to highlight. Not a mystery percentage.

Get a tailored resume highlighting what this role needs.

Role summary

SoFi is seeking an AI Engineer to join its independent risk organization. This hands-on role focuses on building agentic AI systems for high-impact risk management and internal workflow problems. You will design, prototype, and deploy AI systems leveraging LLMs, agents, and orchestration, working at the intersection of intelligence and user experience. Responsibilities include architecting agentic systems with multi-step reasoning and tool use, designing user-facing elements, engineering context for LLMs, productionizing AI services, and building evaluation frameworks. The role requires a Bachelor's/Master's degree, 2-5 years of software development experience with AI applications, and familiarity with LLMs, agentic patterns, cloud platforms, and data handling. Python proficiency is preferred.

Employee Applicant Privacy Notice

Who we are:

Shape a brighter financial future with us.

Together with our members, we’re changing the way people think about and interact with personal finance.

We’re a next-generation financial services company and national bank using innovative, mobile-first technology to help our millions of members reach their goals. The industry is going through an unprecedented transformation, and we’re at the forefront. We’re proud to come to work every day knowing that what we do has a direct impact on people’s lives, with our core values guiding us every step of the way. Join us to invest in yourself, your career, and the financial world.

The role:

SoFi’s AI Engineer is a hands-on engineering role within SoFi’s growing independent risk organization, focused on building agentic AI systems to solve real-world, high-impact problems. This role will be instrumental in designing, prototyping, and deploying AI systems that enhance risk management and internal workflows.

This role sits at the intersection of the intelligence layer, including LLMs, agents, and orchestration, and the experience layer, which focuses on how users interact with and derive value from the AI systems developed. You will work closely with the Senior Manager of AI Engineering within the Risk Analytics group to build systems that are technically strong, intuitive, reliable, and impactful for end users.

What you’ll do:

Architect and Develop Agentic AI Systems: Design, build, and orchestrate AI systems that leverage multi-step reasoning, tool use, and structured workflows, using frameworks such as LangGraph or similar approaches. Incorporate planning, memory, tool integration, and adaptive control flow to enable automated decisioning, risk insights, and internal platforms.

Design the Experience Layer: Work closely with stakeholders to define how users interact with AI systems, including designing intuitive workflows, interfaces, and feedback loops that drive adoption and trust.

Context Engineering and System Design: Structure inputs, outputs, and system context to improve reliability and performance of LLM systems, including prompt design, retrieval strategies, and workflow composition.

Productionize AI Systems: Develop production-grade services and APIs, integrate agents into real systems, and ensure scalability, reliability, and maintainability.

AI Observability and Evaluation: Build tracing, debugging, and evaluation frameworks to understand system behavior and continuously improve agent performance.

Cross-Functional Collaboration: Partner with risk, engineering, and business teams to translate ambiguous problems into working AI systems and deliver measurable outcomes.

Proof of Concepts and Innovation: Identify opportunities to automate workflows using AI, rapidly prototype solutions, and evaluate new tools and approaches by staying up-to-date with latest trends in AI

What you’ll need:

Bachelor’s or Master’s degree in Computer Science, Data Science, Artificial Intelligence, Machine Learning, or a related field.

Two to five years of software development experience, with hands-on experience building and shipping AI-powered applications or workflows.

Experience working with LLMs and building applications using prompting, APIs, or agent frameworks.

Familiarity with agentic patterns such as tool use, multi-step reasoning, or workflow orchestration.

Experience structuring inputs and outputs for LLM systems through context engineering, including prompt design and retrieval-based approaches.

Experience building backend services and APIs, with Python preferred.

Familiarity with cloud platforms such as AWS, Azure, or GCP and modern development practices.

Experience working with data, including structured or unstructured data, and building pipelines for downstream applications.

Understanding of how to evaluate AI systems, including defining success metrics and iterating based on performance.

Strong problem-solving skills and the ability to work through ambiguous problems.

Strong communication and collaboration skills, with the ability to work cross-functionally.

An ownership mindset with a bias for action in building and shipping solutions.

Nice to have:

Experience with frameworks for building agentic applications such as LangGraph.

Experience designing user-facing workflows, internal tools, or applications powered by AI.

Familiarity with observability tools for AI systems such as Langfuse / LangSmith

Exposure to financial services or risk-related use cases.

Experience with frontend technologies such as React for building AI-powered interfaces.

Compensation and Benefits

The base pay range for this role is listed below. Final base pay offer will be determined based on individual factors such as the candidate’s experience, skills, and location.

To view all of our comprehensive and competitive benefits, visit our Benefits at SoFi page!

SoFi provides equal employment opportunities (EEO) to all employees and applicants for employment without regard to race, color, religion (including religious dress and grooming practices), sex (including pregnancy, childbirth and related medical conditions, breastfeeding, and conditions related to breastfeeding), gender, gender identity, gender expression, national origin, ancestry, age (40 or over), physical or medical disability, medical condition, marital status, registered domestic partner status, sexual orientation, genetic information, military and/or veteran status, or any other basis prohibited by applicable state or federal law.

The Company hires the best qualified candidate for the job, without regard to protected characteristics.

Pursuant to the San Francisco Fair Chance Ordinance, we will consider for employment qualified applicants with arrest and conviction records.

New York applicants: Notice of Employee Rights

SoFi is committed to an inclusive culture. As part of this commitment, SoFi offers reasonable accommodations to candidates with physical or mental disabilities. If you need accommodations to participate in the job application or interview process, please let your recruiter know or email accommodations@sofi.com.

Due to insurance coverage issues, we are unable to accommodate remote work from Hawaii or Alaska at this time.

Internal Employees

If you are a current employee, do not apply here - please navigate to our Internal Job Board in Greenhouse to apply to our open roles.

Sample SoFi interview questions

1
Develop a real-time system for managing traffic and rerouting.
system designmedium
2
Reverse Nodes in k-Group Reverse nodes in k-group in a linked list. Input: head = [1,2,3,4,5], k = 3 Output: [3,2,1,4,5] Explanation: The first 3 elements are reversed, while the remaining 2 are left untouched since they don't form a complete group.
codingmedium
3
Clone an Undirected Graph Clone an undirected graph. Input: adjList = [[]] Output: [[]] Explanation: Creates a new, deeply cloned graph containing only one single node with zero connected neighbors.
codingmedium
4
Longest Substring with Two Distinct Characters Find the longest substring with at most two distinct characters. Input: s = "ccaabbb" Output: 5 Explanation: The substring "aabbb" contains exactly two distinct characters and has a maximum length of 5.
codingmedium
5
Diameter of a Binary Tree Find the diameter of a binary tree. Input: root = [1,2] Output: 1 Explanation: The longest path is simply the single edge connecting the root node to its only child.
codingmedium

Ready to apply?

You'll be redirected to SoFi's application page.