Job Title: Excel/Python/Language Model Expert for Excel cleanup

Project Description:

We are producing monthly advisories for farmers, and we have these advisories as 3889 rows in Excel. These advisories reflect seasonal production stages, local weather forecasts, and good agricultural practices. They are currently disaggregated in 8 different columns. We need them merged into one column at the end of the Excel sheet to be called ‘Final Advisories’. We also need them cleaned. Cleaning includes:

• In each row, removing duplicates to avoid redundancy, and only retain information that is actionable, accurate, and relevant

• Removing all references to chemicals
We tried to use ChatGPT, but it has proven to have limitations, for example, occasionally it includes chemical recommendations, even when explicitly instructed not to. That’s why we are seeking human expertise or advanced automation that won’t overlook these critical details.

Who We’re Looking For:
We are open to two types of profiles:
1. Python Developers / AI Engineers with Agronomic Knowledge
You know how to process Excel data programmatically and can build or use a custom language model, script, or ruleset to reliably generate error-free advisories. If you can build logic that ensures no chemical references are included, we want to hear from you.
2. If you have another method that can achieve similar or better results in less time, we also want to hear from you.
Key Responsibilities:
• Review multiple advisory-related columns per row in a provided Excel file.
• Generate one well-structured, user-friendly advisory per row, using:
o Clear, specific language (e.g., instead of “irrigate,” specify how much water, how often, and how to test soil moisture).
o Organic, mechanical, or cultural alternatives (no chemicals).
o Farmer-friendly measurement proxies (e.g., “5–10 cm mulch (about two fingers thick)”).
• Ensure each row is self-contained (not dependent on previous advice).
• Work fast but carefully — we have only 2 days to complete this.
Deliverables:
• The same Excel file returned with an added column titled “Final Advisory”, containing the cleaned text for each row.
• A quality assurance method to confirm the work was done correctly:
o If you're working in Python or another script, please provide the script. We will review your code rather than manually reviewing each row.
o If you’re editing manually, provide a brief description of your checking process or validation steps to ensure completeness and compliance.

Interface

Prompt Setting: