AI Orchestration Engineer

Company Overview

Soros Fund Management LLC (SFM) is a global asset manager and family office founded by George Soros in 1970. With $28 billion in assets under management (AUM), SFM serves as the principal asset manager for the Open Society Foundations, one of the world’s largest charitable foundations dedicated to advancing justice, human rights, and democracy.

Distinct from other investment platforms, SFM thrives on agility, acting decisively when conviction is high and exercising patience when it’s not. With permanent capital, a select group of major clients, and an unconstrained mandate, we invest opportunistically with a long-term view in a wide range of strategies and asset classes, including public and private equity and credit, fixed income, foreign exchange, and alternative assets. Our teams operate with autonomy, while cross-team collaboration strengthens our conviction and empowers us to capitalize on market dislocations.

At SFM, we foster an ownership mindset, encouraging professionals to challenge the status quo, innovate, and take initiative. We prioritize development, enabling team members to push beyond their roles, voice bold ideas, and contribute to our long-term success. This culture of continuous growth and constructive debate fuels innovation and drives efficiencies.

Our impact is measured by both the returns we generate and the values we uphold, from environmental stewardship to social responsibility. Operating as a unified team across geographies and mandates, we remain committed to our mission, ensuring a meaningful, lasting impact.

Headquartered in New York City with offices in Greenwich, Garden City, London, and Dublin, SFM employs 200 professionals.

Team Overview

We are seeking an AI Orchestration Engineer to work closely with portfolio managers, analysts, and researchers to design and optimize AI-driven workflows that enhance decision-making and investment processes. You will scope use cases, configure and test solutions, and iteratively refine them based on stakeholder feedback to deliver clear, actionable insights.

This role will focus on high-impact applications such as market intelligence synthesis and context management for large datasets. Examples projects include tonal analysis of earnings calls or summarization of Bloomberg IB chats. Success requires the ability to translate nuanced research questions into practical workflows that are both effective and seamlessly embedded in daily investment processes.

To excel, you must combine strong problem-solving skills with fluency in modern AI orchestration frameworks and the judgment to select the right approach—whether configuring prompts, tuning retrieval strategies, or leveraging third-party tools. A collaborative mindset and clear technical communication are essential.

Major Responsibilities

• Partner with portfolio managers and analysts to identify opportunities where AI workflows can enhance research, operations, and decision making.

• Configure and adapt orchestration frameworks such as LangChain, LangGraph, Bedrock, and OpenSearch to facilitate advanced orchestration patterns.

• Test problem-specific techniques to circumvent the context window’s token cap, such as sequential chunking or pre-processing RAG data sets.

• Define model guardrails, engineer prompts, and refine context management strategies.

• Optimize OpenSearch settings per data set (text enrichment, pre-processing, vector encoding, and retrieval strategies) to improve performance for nuanced queries and big data.

• Provide detailed documentation for each solution, develop best practices, and share knowledge across teams.

• Stay current with advances in LLM orchestration and RAG techniques.

What We Value

• Bachelor’s degree in computer science, engineering, or related field.

• 2-5 years of relevant experience on the job or in a specialized graduate program.

• Proficiency with LLM orchestration frameworks, especially regarding context management strategies.

• Experience configuring model guardrails, engineering prompts, and fine-tuning.

• Hands-on experience with text enrichment, embedding, and retrieval strategies in vector databases such as OpenSearch.

• Strong Python skills for workflow configuration and automation.

• Excellent communication skills with demonstrated ability to partner with non-technical stakeholders.

We anticipate the base salary of this role to be between $150,000-200,000. In addition to a base salary, the successful candidate will also be eligible to receive a discretionary year-end bonus.

In all respects, candidates need to reflect the following SFM core values:

Smart risk-taking // Owner’s Mindset // Teamwork // Humility // Integrity