I Let AI Apply to 819 Jobs for Me. Here’s What Happened

An experiment in automation, rejection, and what it says about today’s job market

The job market is broken. There’s nothing normal about this current situation, so I decided to embrace change and take a risk.

I used AI agents to auto-apply to 819 jobs in one month. Here’s how it went.

---

---

What I did

I used three AI Agents with three separate job buckets (one Agent for each type).

1. AI-focused roles

2. Leadership/ people management roles

3. Program/ client management roles

I created tailored CVs that laid out my relevant experience. The AI would then identify jobs and tailor my application, CV, and cover letter to align with the job description. I also gave the AIs the information that would allow them to fill in all of those online questions like gender, military, and disability status, as well as the boxes with the motivation for applying, or career aspirations, etc.

I then had two options:

• I could either review every application before allowing the AI to submit it; or

• I could set it to “auto-apply,” and it would do everything automatically.

For the first bit, I took the approach of reviewing the application, which allowed me to train the AI to tailor my resume and cover letter the way I wanted them to. This reduced the risk of the AI misrepresenting my experience.

I then switched it to “auto-apply” and periodically reviewed the applications to make sure they were aligned and accurate.

I had three choices for how broadly to let it apply: stick to jobs I was a strong match for, loosen it slightly to include roles just outside my skills, or open the gates to positions well beyond my experience.

I chose the middle ground.

Results

Approximately $50- and 2-days’ worth of effort setting up and monitoring three AI agents got me five interviews. That is staggering alone, but there were some additional benefits that I had not considered.

1. I felt less guilty. Having an AI apply to an insane amount of jobs felt like I was doing my due diligence, which meant that my brain took some of the weight off, and I was able to think more clearly.

2. I didn’t take rejections personally. Reading through my email was a pleasant journey through rejections from companies I didn’t even know existed. I hadn’t put any emotional energy or time into the application, so it was easy not to take it personally. In fact, sometimes I read them and thought “you right! I’m not the right person for that.”

3. It helped me identify opportunities I otherwise would have missed. In reviewing the applications, I found organizations that I actually like, and was able to use traditional searches to find other opportunities with them (which led to interviews).

However, this experiment had its dangers and drawbacks.

1. There was a risk of misrepresentation that I had to live with. I walk through below how I mitigated it to the best of my ability and greatly reduced the risk, but you can never quite be sure how an AI is going to characterize your experience.

2. I got interviews for positions and organizations that I didn’t want to work for. This included political organizations that I did not feel comfortable with.

Here’s what the numbers looked like (they’re not pretty)

• 819 jobs

• 71 responses (Response rate of 8.6%)

• 5 interviews (6 roles, as one interview was for multiple)

That’s an interview rate of 0.6%

The five interviews are not quite the whole picture.

• Two of the companies I got interviews for were not places that I would ever want to work. I politely declined.

• One interview started the conversation with “we changed our mind and are changing the role.” Fine. I thanked them for their time.

This leaves us with just two companies (and three roles).

Attributing the results to a crashing job market, my CV, or the way the AI tailored the applications is impossible from these data alone. It’s likely a mixture.

Based roughly on these numbers, I could expect one interview for every 200 applications.

Would I have gotten a better response if I had done 819 applications without AI? Yes, probably.

But the additional benefits of doing it this way really shocked me. Particularly, the mental benefits.

Using AI Agents to auto-apply freed me to focus on activities that had a greater overall reward, like connecting with people, exploring new ways to make money, writing thought pieces, or even just taking a mental break.

It had other benefits. Like, I was also not crushed by any of the rejections; in fact, I found the daily rejection email reading kind of fun. It felt like discovering what my little AI elves had been doing while I slept. Often, I had never heard of the companies that were rejecting me, which largely eliminated any offense I could have felt. I read my emails in the morning with a smile and a “I’m sorry, who are you?”

There’s a contrarian voice in my head that says: “but the use of AI Agents for hiring is only contributing to the AI job hiring hellscape, how could you add to that?” And there is truth to it: if even a few people use these AI Agents, they could completely overwhelm hiring, forcing hiring managers to rely completely on AI to assess candidates, which then can exacerbate in-built bias (LLMs prefer their own writing, plus hiring biases have been well documented in AI systems). However, hiring managers are likely to adopt AI more, regardless of what you or I personally do. Almost everyone is already using AI to adjust or write their applications; our choice here is whether we want to utilize the ability to put our names into more hats than would be possible manually.

For me personally, by offloading the repetitive, demoralizing parts of the process, I was able to reclaim mental bandwidth for the things that actually matter: building relationships, exploring opportunities that align with my values, and thinking creatively about current and future survival. In a way, the AI Agents were insulating me from the worst parts of a broken system so I could show up sharper where it counts.

This brings me back to the starting point: nothing about this market feels normal. Taking a small risk with AI felt justified, and in doing so, I found benefits I didn’t expect—quiet, practical gains that made the process more bearable.

If you’re curious about the AI Agent I used, drop a comment and I’ll DM you.

I now help organizations use AI in ways that are useful, safe, and responsible. If you know of a foundation, nonprofit, or company that’s starting to explore AI and could use guidance, please connect me. Just reply, DM me, or pass this along—I’d be grateful for the introduction.