Machine Learning Engineer Intern - Research at GoodAtNumbers in Remote

GoodAtNumbers is building an always-on decision intelligence platform that is trying to replace what a data scientist, data analyst and a business analyst does. We are looking for someone who can help us push both the research quality and production quality of our ML systems forward.

We are hiring a Machine Learning Engineer Intern for a paid 12-week summer internship from May through July 2026. This is a remote role based in the United States and is expected to be 40 hours per week. Compensation for this internship is $30/hour.

This role sits at the intersection of ML research, software engineering, and MLOps. You will work on problems related to retrieval, context construction, model/tool orchestration, evaluation, monitoring, and the productionization of AI systems. This is a strong fit for someone who can move from experiments to production code and who wants to work on real product problems instead of isolated notebooks.

What you’ll work on

• Design and run experiments across areas such as retrieval, ranking, context construction, tool use, grounded generation, model evaluation, anomaly detection, forecasting, or optimization workflows
• Improve the quality, reliability, latency, and observability of ML and LLM-driven features
• Build reproducible evaluation workflows for model behavior, answer quality, grounding, failure analysis, and regression testing
• Help productionize research work through pipelines, APIs, services, monitoring, versioning, and deployment workflows
• Improve MLOps practices around experiment tracking, prompt/model versioning, dataset versioning, testing, rollout safety, and post-deployment monitoring
• Collaborate closely with software and platform engineers to ship ML systems that are useful, measurable, and production-ready

What success looks like by the end of the internship

• At least one meaningful ML or LLM system is measurably improved in quality, reliability, or latency
• Research work is backed by reproducible evaluation and monitoring rather than one-off experimentation
• The path from experiment to production is cleaner, faster, and safer

What we’re looking for

• 3–4 years of relevant experience preferred through research labs, internships, startups, open-source work, or production ML systems
• Strong software engineering ability and strong comfort writing production-quality code
• Strong Python skills preferred
• Experience with machine learning experimentation, evaluation, and debugging preferred
• Experience with LLMs, retrieval systems, vector search, ranking, prompt/tool workflows, or agent-style systems preferred
• Experience with MLOps practices such as experiment tracking, versioning, model testing, deployment, and monitoring preferred
• Comfort with statistics, error analysis, benchmarking, and translating ambiguous research ideas into shippable systems
• Strong communication and the ability to document tradeoffs, assumptions, and results

Nice to have

• Experience with PyTorch, Transformers, or modern ML tooling
• Experience with vector databases, RAG systems, or evaluation harnesses
• Experience with time-series forecasting, causal analysis, anomaly detection, or optimization systems
• Experience with Docker, Kubernetes, cloud infrastructure, or batch/orchestration systems
• Publications, benchmark work, or strong public repos/writeups

Work authorization
Applicants must be authorized to work in the United States for the full internship period and must be based in the U.S. during the internship. We are not able to provide employment visa sponsorship for this internship.

We welcome applicants from all backgrounds and evaluate candidates based on technical depth, execution, communication, and fit for the role.