Getting Started


During onboarding you will be asked for Country, Search keyword and Location.

<img src="/assets/images/01-onboarding.png" alt="Onboarding" width="300" />
*Figure 1: Onboarding page*

You can also update the search parameters on [Match page](/match).

<img src="/assets/images/01-match-edit-search-parameters.png" alt="Job match" width="600" />
*Figure 2: Match page - Edit search parameters*


Job Matcher will do a search using the specified parameters and return a list.
- Search text: More specific search will narrow your options mean less jobs on the list.
- Location: Search will prioritise results near the selected location. This defaults to user's country if not provided.

JobMatcher updates the job list everyday.


Specify search parameters


Sort and filter job results by using criteria.

### Enter criteria

Click edit icon on Matching criteria panel. On each criteria assign a score and the text that will be evaluated by LLM.

<img src="/assets/images/02-edit-criteria.png" alt="Edit criteria" width="600" />
*Figure 1: Enter criteria*

### Evaluate criteria

Click Evaluate button

<img src="/assets/images/02-evaluate-criteria.png" alt="Edit criteria" width="600" />
*Figure 2: Evaluate button*


Each criterion will be evaluated using large language model with Ollama. If criterion matches it will contribute to the overall score of the job. 

<img src="/assets/images/02-evaluating-criteria.png" alt="Edit criteria" width="600" />
*Figure 3: Evaluation in progress*


You can also assign a negative score for a criterion that you like to exclude, in effect will reduce the job's score. In this example we assigned minus 500 to jobs that require 5 day week in the office.

The job list is sorted by score then by recency. This makes most relevant job according to your criteria to be on top.

### Filter jobs

You can also filter the job list by toggling on the criteria chips.

<img src="/assets/images/02-filter-criteria.png" alt="Edit criteria" width="600" />
*Figure 3: Filtered list using criteria*

In the example above we selected jobs that matches Hybrid, Salary more than 90K and in Financial services.


Sort and filter jobs with matching criteria


After specifying your search parameters and evaluation of matching criteria, you will have a sorted list of jobs with most relevant on the top.

### Filter jobs by seen or recency

You can further filter the list by:
- unseen/seen - when you want to toggle between seen or new jobs.
- last day/last week - when you like to see only the recent jobs.

<img src="/assets/images/03-filter-jobs.png" alt="Filter jobs" width="600" />
*Figure 1: Filtered list using Unseen toggle*

### Job card

On each job card you will find:
- Job title
- Company
- Location
- Date found
- Source website
- Criteria badges
- Action buttons

<img src="/assets/images/03-job-card.png" alt="A job card" width="300" />
*Figure 2: A job card*


### Criterion badge

Each badge represents criterion assessed, with it's score. If the badge is green this means the assessment is positive and it adds to overall score.
Click or hover on a badge to see the AI assessment details.

<img src="/assets/images/03-job-card-criterion-badge.png" alt="Job card criteria matches" width="300" />
*Figure 3: Criteria badge displays the AI assessment*

### Job Page

You can see the job posting details by clicking on the title or view button from the job card. When viewed the job post will be marked seen.

<img src="/assets/images/03-job-page.png" alt="Job page" width="600" />
*Figure 4: Job page*

In the job page you will see:
- Job post title, location, source and description full text
- Job criteria matching details
- Action buttons

### Actions

- Save - the job will be marked as `Saved` and will appear on `My job list`.
- View post - this will take you to the source page, and a prompt will appear to ask if you applied. The job will be marked as `To apply` or `Applied` and will appear on `My job list`.
- Remove - this will stop showing the job.

Select jobs to apply


In `My job list` page you will find the jobs that you saved, post viewed or marked applied.

<img src="/assets/images/04-job-list.png" alt="Filter jobs" width="600" />
*Figure 1: My job list*

From here you can update the status of your job application and keep track its progression.

- Interviewing - initial stage: First contact with the recruiter.
- Interviewing - next stages: Job progressed to the following stages.
- Offer received - congrats!

Manage job list

Ollama


# Why use Ollama?

You can save cost by using AI on your machine instead of using pay per use services from LLM providers (Open AI, Google, etc.)
Job matcher talks to Ollama for job criteria analysis. When writing criteria for jobs it may take a couple of text variation before coming up with an effective criteria. You will be able to
experiment with prompts without worrying about the cost.

Ollama supports a variety of models. Our criteria evaluation defaults to Llama 3 but you can configure it to use any model available in Ollama.


Ollama is an open-source platform designed to run large language models (LLMs)  locally on personal computers or servers. It aims to make AI technology more accessible  by allowing users to download, run, and customize various AI models without relying on cloud services.

What is Ollama?


## Download and Install

Download Ollama from the official page: [https://ollama.com/download](https://ollama.com/download)

The installation process varies slightly depending on your operating system.

## Choosing model

Llama3 is a good model to use as baseline. If you have more capable machine we recommend models from OpenAI like gpt-oss-20b.

To use llama3, run this on terminal:
```
  ollama pull llama3
  ollama run llama3
```

To use OpenAI model:
```
  ollama pull gpt-oss-20b
  ollama run gpt-oss-20b
```

Change the model to use in the Criteria setting on Match page.


## Allow Job Matcher to access Ollama
By default local Ollama instance is not accessible from external websites for security reasons.
We need to explicitly allow Job Matcher to talk to our local Ollama.

### MacOS

Run the following command:

```
launchctl setenv OLLAMA_ORIGINS "https://job-matcher.senyas.io"
```

Restart the Ollama application to apply the changes.

### Linux

For Linux users running OLLAMA as a systemd service, the `systemctl` command is used to set environment variables:

Use `systemctl edit ollama.service` to open the service file in an editor.
In the `[Service]` section, add the Environment line with your CORS settings.

```
[Service]
Environment="OLLAMA_ORIGINS=https://job-matcher.senyas.io"
```

Save your changes, then reload systemd and restart OLLAMA with

```
systemctl daemon-reload
systemctl restart ollama
```


### Windows

Ensure Ollama is not running by quitting the application from the taskbar.
Open the Control Panel and navigate to “Edit system environment variables.”
Choose to edit or create a new variable named `OLLAMA_ORIGINS`.

```
Variable name: OLLAMA_ORIGINS
Variable value: https://job-matcher.senyas.io
```

Press OK and close the control panel.
Run Ollama from a new terminal window to ensure it picks up the updated environment variables.



Install and configure Ollama to work with Job matcher

Setup Ollama

"use strict";
const {Fragment: _Fragment, jsx: _jsx, jsxs: _jsxs} = arguments[0];
function _createMdxContent(props) {
  const _components = {
    code: "code",
    em: "em",
    li: "li",
    p: "p",
    ul: "ul",
    ...props.components
  };
  return _jsxs(_Fragment, {
    children: [_jsxs(_components.p, {
      children: ["In ", _jsx(_components.code, {
        children: "My job list"
      }), " page you will find the jobs that you saved, post viewed or marked applied."]
    }), "\n", _jsx("img", {
      src: "/assets/images/04-job-list.png",
      alt: "Filter jobs",
      width: "600"
    }), "\n", _jsx(_components.p, {
      children: _jsx(_components.em, {
        children: "Figure 1: My job list"
      })
    }), "\n", _jsx(_components.p, {
      children: "From here you can update the status of your job application and keep track its progression."
    }), "\n", _jsxs(_components.ul, {
      children: ["\n", _jsx(_components.li, {
        children: "Interviewing - initial stage: First contact with the recruiter."
      }), "\n", _jsx(_components.li, {
        children: "Interviewing - next stages: Job progressed to the following stages."
      }), "\n", _jsx(_components.li, {
        children: "Offer received - congrats!"
      }), "\n"]
    })]
  });
}
function MDXContent(props = {}) {
  const {wrapper: MDXLayout} = props.components || ({});
  return MDXLayout ? _jsx(MDXLayout, {
    ...props,
    children: _jsx(_createMdxContent, {
      ...props
    })
  }) : _createMdxContent(props);
}
return {
  default: MDXContent
};
