Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for chemical engineers

Catalyst Performance Tracking Agent

A reliable projected change-out date updated every week

AnalysisForecastingMonitoringScience and researchOperations
See how it worksGet access
Catalyst Performance Tracking Agent: what goes in, what the agent does and what you get

What it does

Catalysts lose activity slowly, and planning a change-out needs months of notice for ordering and shutdown. Each week this agent pulls activity, selectivity, inlet temperature and pressure drop from plant and lab data. It removes upset and shutdown periods and normalizes the rest for rate, feed and temperature so trends are fair to compare. It fits a decline curve and projects when the bed will reach your change-out limit. If the decline suddenly speeds up, it checks feed analyses for poisons and the event log for upsets before refitting. It then checks whether the projected date falls inside the ordering lead time and warns if so. The engineer reviews the projection and decides timing and orders. Edge case: data from a short upset is left out so it does not distort the projection.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Weekly catalyst update 2 USES A TOOL Pull plant and lab data 3 DOES Remove upset periods and normalize for conditions 4 USES A TOOL Fit the decline trend and project the limit date 5 CHECKS THE RESULT Is the decline rate consistent with recent weeks? If not: check feed analyses and event log, then refit.Back to step 3. 6 CHECKS THE RESULT Is the projected date outside the ordering leadtime? If not: warn the engineer and draft an early order note.Back to step 4. 7 DOES Write the catalyst status report 8 YOU APPROVE Engineer reviews the projection 9 RESULT Catalyst status report
Read the steps as a list
  1. Weekly catalyst update
  2. Pull plant and lab data
  3. Remove upset periods and normalize for conditions
  4. Fit the decline trend and project the limit date
  5. Is the decline rate consistent with recent weeks?If not: check feed analyses and event log, then refit. Back to step 3.
  6. Is the projected date outside the ordering lead time?If not: warn the engineer and draft an early order note. Back to step 4.
  7. Write the catalyst status report
  8. Engineer reviews the projectionThe agent waits here for your OK.
  9. Catalyst status report

How it decides

It fits a decline trend on normalized data and flags when the rate of decline changes beyond the set amount.

  • Exclude upsets and shutdown periods
  • Flag decline rate changes above the threshold
  • Warn when the projected date is inside the ordering lead time

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Change-out limits
  • Normalization method
  • Decline rate threshold
  • Ordering lead time (default 16 weeks)

What keeps you in control

It always asks you first

  • Change-out timing
  • Catalyst orders

Hard limits

  • Never schedules shutdowns or places orders

It stops when

  • Done: report filed
  • Stop: key data missing; ask the lab

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensAfter week 30, activity started falling twice as fast as before, so the consistency check failed. The agent checked feed analyses and found sulfur rising from 0.5 to 1.4 ppm. It flagged likely poisoning, refit from week 30 and moved the projected change-out from March to January. The lead time check then failed against a 16-week order time, so it warned the engineer, who placed the order.

More agents for chemical engineers

Analysis · Decisions

Process Parameter Window Study Agent

Define an operating window that the data actually support, with extra runs where it is thin.

11 steps2 checks7 jobs
Analysis · Science and research

Raw Material Variability Study Agent

Identify which raw material attributes drive process outcomes and propose tighter specifications backed by data.

10 steps2 checks6 jobs
Compliance · Science and research

Hazard Study Recommendation Closure Agent

Open hazard study recommendations are tracked to documented closure or a justified decision.

11 steps2 checks1 job
Monitoring · Compliance

Safety Device Test Schedule Agent

No safety device goes past its test date, and every failed test is followed to a passing retest.

11 steps3 checks5 jobs