Research Engineer for Financial Reasoning (LLMs)

---

About Deep Insight Labs

Deep Insight Labs is an applied AI research company building agentic AI systems for investment research and financial reasoning. Our flagship platform, Vector, combines large language models, structured knowledge representations, and human-in-the-loop reasoning to solve complex, real-world analytical problems faced by investors and financial professionals.

We operate at the intersection of cutting-edge AI research and production-grade systems, with active efforts to publish in top-tier ML conferences and journals while deploying novel models into real products.

Internship Overview

We are seeking a highly motivated Master’s student in Computer Science (or related field) for a research and development internship. The intern will work on novel methods at the intersection of knowledge graphs, representation learning, and LLM-based reasoning, with joint supervision from Deep Insight Labs and Zhongtian Sun, Assistant Professor in AI (MILA, Cambridge, Oxford, University of Kent)

This internship is designed for candidates who are serious about research excellence and want to see their work both published and deployed.

Research Focus

You will work on one or more of the following areas:

• Knowledge Graph Construction & Learning
  • Financial knowledge graph modeling (entities, events, relationships)
  • Graph neural networks (GNNs) and representation learning
  • Temporal and heterogeneous graphs
     
• Representation Learning
  • Learning structured representations from unstructured financial text
  • Joint embedding of graphs, documents, and numerical signals
  • Contrastive and self-supervised learning approaches
     
• Financial Reasoning with LLMs
  • Multi-step reasoning and planning with LLMs
  • Integrating symbolic structures (graphs, rules) with neural models
  • Explainability, attribution, and reasoning traceability
     
• Research → Production
  • Translating research ideas into production-ready models
  • Evaluating models under real-world constraints (latency, robustness)
  • Building experimental pipelines and benchmarks
     

Expected Outcomes

• Submission to a top-tier venue (e.g. NeurIPS, ICML, ICLR, KDD, WWW, or leading journals)
• Production deployment of research ideas into Deep Insight Labs’ platform
• Open-source contributions (where appropriate)
   

Who Should Apply

We are looking for candidates who:

• Are currently pursuing a Master’s degree/PhD in Computer Science, AI, or related field
• Have strong foundations in machine learning and deep learning
• Are comfortable with Python and at least one ML framework (PyTorch preferred)
• Have prior exposure to NLP, graphs, or representation learning (coursework or projects)
• Are interested in bridging theory and practice
• Can work independently and communicate research ideas clearly
   

Nice to Have (but not required):

• Experience with knowledge graphs or GNNs
• Familiarity with LLMs and prompting / fine-tuning techniques
• Prior research experience or publications
• Interest in finance or economic reasoning
   

What We Offer

• Joint academic–industry supervision
• Direct mentorship on research framing, experimentation, and writing
• Opportunity to co-author a top-tier publication
• Experience deploying novel AI models into production
• Access to compute resources and proprietary datasets
• A fast-moving, research-driven environment with real-world impact
   

Internship Period & Mode of Work

Mar to Sep 2026 (Flexible), Full Remote