AI for Environmental Scientist / Sustainability Analyst
An environmental scientist spends an estimated 18 of 45 hours in an office week on data review, tables, figures, records and report drafts, and in a field-heavy week an estimated 22.5 of 45 hours go to field work, field records and sample custody. The guides below show where an AI tool can draft, organize, explain and check around that work, from a field day checklist built from your approved work plan to a weekly digest of lab results that are late. The lab reports the numbers, your spreadsheet does the comparing, and your project manager owns every conclusion.
Try right now
Copy a prompt, paste into ChatGPT, Claude, or Gemini
Works with any free AI chatbot, no signup needed
A rewritten version of your request that asks for what an experienced professional would ask for.
Here's a task I'm about to do: [describe the task]. Don't do it yet. First, tell me how the best person in my field would approach it: what they would check that I haven't mentioned, what raw information they would want instead of a summary, and how they would prove the result is right. Then rewrite my request so it asks for that version.
View full prompt →Tip: - Check the result against something outside the chat: the official source, last period's actual figures, a colleague who knows the area, or a calculator. The AI saying its answer is right proves nothing.
A one- or two-paragraph site description in report style that uses only your notes and flags every missing detail with "[confirm]".
Draft a site description section of one or two paragraphs in plain report style, using only my notes below. Mark anything missing "[confirm]". State no past use, regulatory status or site condition that my notes do not give. Label it a draft for my project manager's review. Notes: [paste your site notes]
View full prompt →Tip: Past use and regulatory status are the places a chatbot is most likely to invent something plausible, so treat any sentence about them as an error unless you wrote it in your notes. Use coded notes (Site S-01) in place of the site address, owner or tenant names, and keep the questionnaire, site photos and boring logs in your firm's approved systems. Photos carry location data, so strip it before a photo leaves those systems. Check the tool's data settings for training and retention first, and for real project notes confirm that the firm and the client allow the tool.
A sample ID pattern with worked examples, a tracking table layout for sample custody, and a list of the fields you must fill in from the custody form.
Write a sample ID pattern [Site code]-[Location]-[YYMMDD], with -DUP added for duplicates and TB-[YYMMDD] for trip blanks. Give examples. Then lay out a tracking table with columns Sample ID, Location, Date, Time, Matrix, Analyses requested, Cooler, Shipped date, Lab received date, Lab sample ID, Notes. List the fields I must copy from the custody form. Do not decide what the lab or my firm requires.
View full prompt →Tip: Your work plan or your firm's procedure may already set an ID pattern, and the custody form belongs to the lab, so check both before you adopt anything the AI proposes. Run each ID in the example through the pattern by hand to catch a swapped date or location. The prompt uses only codes and column names, which is all the AI needs. Keep real client names, addresses and lab reports out of the chat and in your firm's approved systems, and check the tool's data settings on training and retention.
Spreadsheet formulas for a comparison column, outcome counts, an average of detected results and a duplicate relative percent difference, each with a one-line explanation.
My Results sheet has columns A Location, B Sample date, C Result, D Qualifier (ND or blank), E Screening value, F Comparison. Write the F2 formula to fill down, a COUNTIF for each outcome, an average of detected results, and a relative percent difference for a duplicate pair. Explain each in one line. Do not calculate any results.
View full prompt →Tip: The spreadsheet does the math and you check every result. Whether non-detects count in an average, and how, is your project manager's call, so this example leaves the ND row out only to show the formula. Use column names and invented values in the chat. Real lab results stay in your firm's approved systems, and the tool's data settings on training and retention are worth checking before you paste a live sheet.
A three-tab spreadsheet layout with column names and dropdown values for each tab, ready to build in Excel or Google Sheets.
Design a project tracker spreadsheet for [my team] with three tabs: Field events, Lab batches, Deliverables. For each tab, give a table of column names and the dropdown values for status columns. The only formula is a days-until column: =IF(C2="","",C2-TODAY()). Put the date it counts to in column C.
View full prompt →Tip: The tracker is for your own planning. The Level 4 digests use their own spreadsheets, so do not wire them into this one. Keep the dates in column C real dates, not text, or the formula returns an error. The sheet holds site codes, not site names, so keep client and site identities out of it and out of the chat. The prompt itself needs only column names, but check the tool's data settings on training and retention before pasting anything from a live tracker.
A draft of about 200 words at a general-public reading level that explains a screening value, repeats your counts exactly as given and says what happens next only as your project manager approved.
Write a plain-language summary of about 200 words for [audience] at a general-public reading level. Explain in general terms what a screening value is. Repeat only these counts: [counts]. Say what happens next only as stated in these approved messages: [messages]. Never call the water or site safe or unsafe. End with "Draft for project manager review".
View full prompt →Tip: The explanation of a screening value is the risky sentence, because a chatbot can easily imply that a value separates safe from unsafe. Compare it with your project manager's wording and delete anything that goes further. The counts must come from your spreadsheet, and the AI should never recompute them. Use the coded counts and the approved messages only. Client names, meeting attendee details and unpublished reports stay in your firm's approved systems, and the tool's data settings and the firm's and client's approval still apply.
A short, polite email to your lab contact covering the cooler shipment, a bottle order for the next event and a request to confirm timing.
Draft an email of under 120 words to [lab contact] at [lab]. Say a cooler with [number] samples ships on [date]. Ask for a bottle order for our next event on [date]. Ask them to confirm the holding times and the results due date for this batch. State no holding time or turnaround yourself. Plain, polite tone.
View full prompt →Tip: The lab sets its own holding times and turnaround, so the prompt asks the lab to confirm them and the AI should add no numbers of its own. Delete any it adds. Leave out the client name, the site address, sample results and your contacts' phone numbers. Codes and counts are enough for the chatbot. Check the tool's data settings on training and retention, and if the email needs more project detail, write that part yourself in your firm's approved email.
A caption for each figure with number, title, date and data source line, plus a figure list ready to paste into the report front matter.
Write a caption for each figure below with figure number, title, date and a data source line. Describe only what is plotted. Never interpret it. Use only my descriptions and mark anything missing "[confirm]". Then write a figure list. Label the output a draft for review. Figures: [paste figure descriptions]
View full prompt →Tip: The words "posted values of Analyte A" describe the figure, while "shows contamination" would be a conclusion that belongs to your project manager and the licensed professional, so cut any caption that drifts toward interpretation. Add the real report number and dates yourself from the lab report. Descriptions made of figure titles and codes are enough for the AI. The figures, base maps and lab reports stay in your firm's approved systems, so check the tool's data settings before you paste anything beyond that.
A list of what could go wrong with your plan, each with a prevention check and who to call, plus a ranking of the three that would cost you the day.
Run a pre-mortem on this field day plan. Cover access, weather, a dry well, a broken bottle, a missed duplicate, cooler temperature and courier cutoff. For each, write a prevention check and who I should call, as questions. Give no safety instructions. Then rank the three most likely to cost me the day. Plan: [paste day plan]
View full prompt →Tip: Use the answers as a checklist to talk through with your project manager. The health and safety plan stays the authority for any hazard, so ignore any item that sounds like a safety instruction and take it to your health and safety lead. Paste a coded plan only (Site S-01, Lab A). The site address, access instructions and client contacts stay in your firm's approved systems. Check the tool's data settings on training and retention first.
A day plan in sampling order, a bottle and equipment checklist and a confirm-before-you-go list, all built from the work plan summary you paste and labelled as a draft for your project manager.
Using only the work plan summary below, draft a field day plan in sampling order, a bottle and equipment checklist, and a confirm-before-you-go list. Take bottle types, preservation and holding times only from the summary or the lab's bottle order. Write "check with the lab" where they are missing. Label it a draft for my project manager's review. Summary: [paste work plan summary]
View full prompt →Tip: The prompt works because it fences the AI inside your summary, so a gap shows up as "check with the lab" instead of a plausible-looking bottle type. Compare the finished checklist line by line with the lab's bottle order and the approved work plan, since the AI has seen neither. Paste a summary built from codes (Site S-01, MW-3, Lab A) and not the full work plan, which carries the client name, site address and access instructions. Those documents belong in your firm's approved systems. Before you paste anything real, check the tool's data settings for chat training and retention, and confirm that the firm and the client allow that tool for project material.
A field log table in a fixed column order, with every gap in your notes marked and a short list of questions for you to answer from memory or the field book.
Turn my rough field notes into a daily field log table with columns: time, location, activity, observations, samples collected, deviations. Use only what is in the notes. Mark every gap "[not in notes]". Do not add or interpret anything, including odor or color. Then list one question for each gap. Notes: [paste rough notes]
View full prompt →Tip: Keep the "do not interpret" line, because a chatbot will happily turn "slight odor" into a finding, and descriptions of odor or color belong to you. If a gap is a real deviation from the work plan, that call is yours and your project manager's, not the AI's. Your field notes may hold site names, owner names, addresses and photos. Paste only coded versions (MW-2, Site S-01). The real notes and any photos stay in your firm's approved systems, and a photo loses its location data before it goes anywhere outside them. Check whether the tool trains on your chats and how long it keeps them, and confirm the firm and client allow it before you paste real notes.
Matched pairs, any sample missing from the lab's list, any ID that differs, and a draft question to send the lab.
Compare these two lists of sample IDs by exact text. Report matched pairs, IDs on my custody list missing from the lab list, and IDs on the lab list that differ from mine. Do not guess which ID was meant. Then draft a short question to the lab. Custody list: [paste] Lab login list: [paste]
View full prompt →Tip: Do not rely on the chat for the check. Put both lists in a sheet, with yours in column A and the lab's in column B, and use =COUNTIF(B:B,A2) beside each custody ID. A 0 means no exact match. Compare the sheet with the chat's answer. Then ask the lab. Real custody forms carry client and site details, so paste only the sample IDs (codes) and keep the forms in your firm's approved systems. Check the tool's data settings on training and retention.
Use AI in your tools
AI features built into tools you already have
No new subscriptions, just features you may not have noticed
Set up an AI assistant
Step-by-step guides for dedicated AI tools
10 to 30 minute setup, then ongoing time savings
Go further
Advanced workflows, automation, and custom AI setups
For when you’re ready to connect tools and automate
Recommended Tools
6Ranked by relevance for environmental scientist / sustainability analyst
- 1
ChatGPT
Field Day Plan and Bottle and Equipment Checklist from Your Approved Work Plan (M1), Sample ID Scheme and Chain-of-Custody Tracking Table Layout (M3) + 1 more
BeginnerVerified Oct 2026 - 2
Claude
Rough Field Notes Cleaned into a Consistent Daily Field Log (M2), Claude Project with Your Report Template Headings, Style Notes and Two Public Reports (M6) + 2 more
BeginnerVerified Oct 2026 - 3
Microsoft Excel
Screening-Comparison Formulas Drafted and Explained in Copilot in Excel (M4)
IntermediateVerified Oct 2026 - 4
Microsoft Word
First Draft of a Report Section in Copilot in Word from Your Notes and Tables (M5)
BeginnerVerified Oct 2026 - 5
Gemini Notebook
Gemini Notebook over Public Agency Documents for a Site Records Review (M8)
IntermediateVerified Oct 2026 - 6
Zapier
Daily Lab Results Watch with Zapier (M10)
IntermediateVerified Oct 2026
Common questions
- What is the best AI tool for an environmental scientist / sustainability analyst?
- 1. ChatGPT: Field Day Plan and Bottle and Equipment Checklist from Your Approved Work Plan (M1), Sample ID Scheme and Chain-of-Custody Tracking Table Layout (M3) + 1 more. 2. Claude: Rough Field Notes Cleaned into a Consistent Daily Field Log (M2), Claude Project with Your Report Template Headings, Style Notes and Two Public Reports (M6) + 2 more. 3. Microsoft Excel: Screening-Comparison Formulas Drafted and Explained in Copilot in Excel (M4).
- How can an environmental scientist / sustainability analyst use ChatGPT or another AI chatbot?
- Start with copy-paste prompts that work in any free chatbot. For example: Matched pairs, any sample missing from the lab's list, any ID that differs, and a draft question to send the lab. A field log table in a fixed column order, with every gap in your notes marked and a short list of questions for you to answer from memory or the field book. A day plan in sampling order, a bottle and equipment checklist and a confirm-before-you-go list, all built from the work plan summary you paste and labelled as a draft for your project manager.
- Do I need technical skills to start?
- No. Level 1 prompts work in any free AI chatbot with no signup beyond the chatbot itself: copy the prompt, fill in the bracketed details, and paste it in. Later levels add AI features in tools you already use, then dedicated AI tools and automation.
New to AI?
The Big Four AI Platforms
ChatGPT, Claude, Gemini, and Grok do roughly the same thing. Pick one and start.
The AI Usage Ladder
See how you use AI today, where that sits on the map, and the one next step to try.
Aim Higher with AI
Doing the same work faster is the smallest win. Three ways to ask AI for more, and why most of us under-ask.
How to Keep Up with AI
The landscape changes fast. A low-effort system to stay informed without drowning.
We update this guide when the tools change. See what's changed →