Can AI make a job search better? My open career journal

Looking for a good technical role is its own kind of research project. You have to discover opportunities, determine whether they are credible, compare their requirements with your actual experience, and explain your work clearly to people who may have very different backgrounds.

I'm approaching this process as an experiment: can a structured workflow, supported by AI tools, make a job search more focused, more efficient, and ultimately more useful?

My background is in civil and environmental engineering, water resources, numerical modeling, and technical problem-solving. I'm interested in roles where that experience matters, including engineering analysis, hydrology and hydraulics, GIS, computational modeling, and adjacent research-oriented work.

This journal will document the process rather than pretend that every application leads to an interview.

The workflow I'm testing

My daily process will aim to do five things:

1. Find credible opportunities. Start with employer career pages and reliable listings, then identify roles that are actually open and relevant.
2. Assess fit. Compare the job description against my real skills, experience, location constraints, and interests rather than applying indiscriminately.
3. Prepare carefully. Use AI to help analyze descriptions, draft or refine application materials, and organize follow-ups, while checking every claim and revising the final wording myself.
4. Apply intentionally. Submit applications only after reviewing their accuracy and deciding that the opportunity is worth pursuing.
5. Learn from results. Record responses, interviews, rejections, and unanswered applications so I can adjust the process based on evidence.

AI agents can speed up tedious steps, but they can also misunderstand requirements, find outdated postings, or produce confident yet inaccurate wording. That is why verification is part of the experiment, not an optional extra.

What will be public—and what won't

I'd like to share aggregate results such as relevant roles identified, applications submitted, replies received, time spent, and lessons learned about different search methods. I may also share reusable scripts, evaluation templates, or examples with personal information removed.

I won't publish private recruiter messages, sensitive contact details, application credentials, or documents that aren't mine to disclose. The goal is to make the method useful to readers without turning a personal job search into a privacy risk.

What comes next

I'll begin with a simple tracking system that distinguishes jobs found from jobs genuinely worth applying to. Then I'll establish a repeatable daily routine and compare what the tools suggest with what I can verify myself.

There may be productive days, disappointing days, and days when the most important discovery is that a job isn't a good fit. All three belong in the record.

Welcome to the AI-assisted career journal.