AI 求職アシスタントで求人を選別し、履歴書を調整する方法

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

## Project: Internal knowledge-base search system

- Role: backend developer
- Time: 2025.03 - 2025.09
- Tech: Python, FastAPI, PostgreSQL, vector search, Docker
- Work:
  - designed document parsing and ingestion
  - added permission checks
  - optimized retrieval latency
- Results:
  - average query latency dropped from 3.2s to 1.1s
  - used by 5 business teams

1
2
3
4
5
6

## Application Preferences

- Prefer: remote/hybrid, AI applications, platform tools
- Acceptable: SaaS, internal systems, data platforms
- Avoid: outsourcing, long-term on-site roles, sales-heavy roles
- Weekly limit: 15 high-quality applications

 1
 2
 3
 4
 5
 6
 7
 8
 9
10

Please extract structured information from this job description:

1. Must-have requirements
2. Nice-to-have requirements
3. Actual responsibilities
4. Implied capabilities
5. Likely interview focus
6. Keywords my resume should address

Do not evaluate my fit yet.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16

Based on my profile and experience library, evaluate this role.

Output a table with:
- job requirement
- my supporting evidence
- match level: strong / medium / weak / none
- whether more evidence is needed
- how to reflect it in the resume

Finally give an application recommendation:
A: strongly apply
B: apply with tailored resume
C: pause
D: do not apply

If there is no evidence, write "no evidence". Do not invent experience.

1
2
3
4
5
6
7
8

Based on this JD, rewrite only the resume sections related to the role.

Rules:
1. Do not add experience I did not have.
2. Do not change project facts.
3. Strengthen relevant keywords.
4. Prefer action + method/tech + result.
5. Output before / after / reason.

1

Participated in internal knowledge-base development, responsible for backend APIs and data processing.

1

Built backend components for an internal knowledge-base search system, including document parsing, permission checks, and retrieval APIs with FastAPI and PostgreSQL; reduced average query latency from 3.2s to 1.1s for cross-team knowledge lookup.

1
2
3
4
5

求人が求める能力
似た経験をしたプロジェクト
自分がしたこと
出た結果
この求人との関係

1
2
3
4
5
6
7
8

Check whether this tailored resume has a clear evidence chain.

For each project, output:
- related job requirement
- whether current evidence is enough
- vague wording
- missing data or context
- anything that sounds exaggerated

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

Based on the JD and my tailored resume, write a short cover letter.

Structure:
1. why I am interested in this role
2. which two experiences match best
3. what problem I can help the team solve

Rules:
- under 300 words
- do not exaggerate
- avoid empty corporate phrases
- sound like a real candidate

1
2
3
4
5
6
7

Review my application log:

1. Which roles got more responses?
2. Which keywords appear in high-match roles?
3. Which gaps repeat?
4. What should I prioritize next week?
5. What small resume changes should I make?

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

You are my AI job assistant. Based on my profile, experience library, and the job description, help me screen this role and tailor my resume.

Output:

1. JD breakdown
2. Matching matrix
3. Application recommendation
4. Resume rewrite
5. Risk check

Important:
- write "no evidence" where needed
- do not fabricate experience, data, company names, or responsibilities
- preserve my own voice where possible