Research Integrity Intern

Agrocorp International is exploring the use of AI-assisted tools to support internal research and analysis workflows across commodity markets. As part of this initiative, we are looking for a detail-oriented **Research Integrity intern** to assist in evaluating the quality and accuracy of AI-generated research outputs. This is a quality assurance and fact-checking role. You will not be building AI systems — you will be stress-testing their outputs against real-world data, with the goal of identifying where and why they fail. **WHAT YOU WILL DO** **Primary Responsibility** * Source-verify every factual claim against primary sources — exchange data, central bank publications, macro data releases, and price feeds. * Classify discrepancies by type: incorrect data, outdated information, calculation errors, or misleading framing. * Maintain a structured audit log documenting each finding, its source, and its potential impact on downstream decisions. * Escalate material errors promptly, as outputs may inform time-sensitive business decisions. **Secondary Responsibility** After an initial audit period, you will contribute to a summary findings report identifying recurring error patterns, high-risk output categories, and recommendations for improving output reliability. **WHAT WE ARE LOOKING FOR** **Required** * Foundational understanding of financial or commodity markets — enough to assess whether a data point is plausible. * Comfort with basic financial arithmetic: percentage changes, spreads, and unit conversions. * Methodical research habits — ability to locate and cite primary sources rather than rely on secondary summaries. **Preferred** * Exposure to commodity or FX markets (agricultural products, energy, or major currency pairs). * Basic Python proficiency — sufficient to run scripts and record outputs in structured formats. * Familiarity with macro data calendars (central bank meetings, economic data releases). **Not Required** * Prior experience with AI or machine learning. * Derivatives or options knowledge.