Optimization Formulation using Natural Language & LLMs
Data Science Institute @ Columbia University
Eligibility
Graduate Only
Accepts Applications Until
Aug 15, 2025
Project Duration
Summer
Description
We are looking for a highly motivated Summer Research Intern to join an exciting project at the intersection of optimization, natural language processing, and machine learning. This opportunity involves close collaboration with Prof. Garud Iyengar and offers the potential for long-term impact and product development.
The goal of the internship is to develop a proof-of-concept system that can:
Interpret natural language descriptions of optimization problems. Automatically formulate these problems. Solve them using industry-standard optimizers such as Gurobi, CVXPY, etc. The prototype will be developed primarily in Python.
Ideal Candidate Profile
We're looking for candidates who are eager to learn, build, and iterate quickly. You should be comfortable working independently and motivated to take ownership of the project.
Required Skills
Required Skills:
Completion of a graduate-level course in optimization (IEOR 4004, 4007, or equivalent). Proficiency in Python programming.
Familiarity with cloud computing platforms (e.g., AWS SageMaker, Bedrock, S3). Basic understanding of large language models (LLMs) and their APIs.
Nice-to-Have Skills:
Experience with data analytics tools (e.g., Pandas, Scipy).
Exposure to machine learning libraries. Prior work with LLM integration in Python workflows.
Additional Information
Duration: Summer 2025 (2–3 months)
Compensation: To be determined based on the skill level
Potential for extended collaboration beyond the internship period.
Note: We do not expect you to have mastered all of these areas, but you should be capable of learning new tools quickly and delivering functional code within 2 months.