Node Types - Jobo Enterprise

Input Node

The entry point for data. Receives the job document from the Outbound Feed sync engine. Every pipeline starts with exactly one Input node — it has no configuration.

The job document

Every job enters the pipeline as a snake_case JSON document. The editor’s sample job button loads a real job from your feed in exactly this shape:

{
  "id": "7d9e1c2a-…",
  "title": "Senior Software Engineer",
  "description": "We are looking for…",
  "listing_url": "https://example.com/jobs/123",
  "apply_url": "https://example.com/jobs/123/apply",
  "source": "greenhouse",
  "source_id": "123",
  "created_at": "2025-01-15T10:00:00Z",
  "updated_at": "2025-01-16T10:00:00Z",
  "date_posted": "2025-01-14T00:00:00Z",
  "valid_through": "2025-03-14T00:00:00Z",
  "is_remote": true,
  "employment_type": "Full-time",
  "workplace_type": "Remote",
  "experience_level": "Senior",
  "responsibilities": ["…"],
  "must_have_requirements": ["…"],
  "preferred_requirements": ["…"],
  "all_skills": ["python", "aws"],
  "hard_skills": ["python", "aws"],
  "soft_skills": ["communication"],
  "benefits": ["…"],
  "education_requirements": ["…"],
  "compensation_min_usd_annual": 120000,
  "compensation_max_usd_annual": 180000,
  "company": { "name": "Acme Corp", "id": "acme-corp" },
  "compensation": { "min": 120000, "max": 180000, "currency": "USD", "period": "yearly" },
  "locations": [
    {
      "display_name": "San Francisco, CA",
      "city": "San Francisco",
      "state": "CA",
      "country": "US",
      "latitude": 37.7749,
      "longitude": -122.4194
    }
  ]
}

Fields that aren’t available for a job (e.g. compensation) are null or absent — design your nodes to tolerate sparse documents.

Filter

Decides whether a job continues through the pipeline. A non-matching job is dropped: it’s skipped for this sync and counted as filtered in the push log — this is not an error. Matching jobs pass through unchanged.

Configuration

Property	Type	Default	Description
`mode`	enum	`include`	include — keep jobs that match. exclude — drop jobs that match.
`logic`	enum	`all`	all — every condition must match. any — at least one must match.
`conditions`	array	—	One or more `{ path, operator, value }` conditions.

Field paths

Conditions reference fields with dot paths, including array indexes: title, company.name, locations[0].country, compensation.min.

Operators

Operator	Matches when…
`equals` / `not_equals`	Value equals the field (case-insensitive for strings; numeric and boolean aware)
`contains` / `not_contains`	Strings: substring match. Arrays: any element equals the value.
`greater_than` / `greater_than_or_equal`	Numeric or date comparison
`less_than` / `less_than_or_equal`	Numeric or date comparison
`exists` / `not_exists`	The field is present and not `null` (no value needed)
`in`	Field equals any entry of a comma-separated list (`US, GB, NL`)
`matches_regex`	Case-insensitive regular expression match

Example: remote jobs in the US or UK

Mode: include, Logic: all
is_remote equals true
locations[0].country in US, GB

Place Filter nodes first — right after Input. Everything downstream (especially AI nodes) only runs for jobs that pass, which directly cuts LLM cost.

Field Mapper

Renames and restructures fields with simple source → target path mappings — the no-code alternative to a Liquid template for schema mapping.

Configuration

Property	Type	Default	Description
`mode`	enum	`passthrough_plus_mapped`	mapped_only — output contains only the mapped fields. passthrough_plus_mapped — output is the input document plus mapped fields, minus dropped paths.
`mappings`	array	—	`{ source_path, target_path }` pairs. Nested target paths create intermediate objects.
`drops`	array	—	Paths removed from the output (passthrough mode only).

Behavior

A mapping whose source path doesn’t resolve is skipped silently — sparse documents are normal, not errors.
Values are copied as-is: objects and arrays can be mapped whole (locations → offices).
Nested targets like job.details.title create the intermediate objects automatically.

Example: flatten to a custom schema

Mode: mapped_only

Source	Target
`id`	`id`
`title`	`job_title`
`company.name`	`company_name`
`listing_url`	`url`
`locations[0].country`	`country`
`date_posted`	`posted_at`

Liquid Template

Reshape and remap JSON using Liquid template syntax for cases the Field Mapper can’t express — conditionals, loops, string formatting. The node renders its template against the parent node’s output. The template should produce valid JSON; plain-text output is automatically wrapped as { "result": "<text>" }.

Property	Type	Required	Description
`template`	string	Yes	Liquid template string that outputs JSON

Available Variables

The input document is bound to data — access any field of the job document with dot notation: data.title, data.company.name, data.locations[0].city.

Available Filters

All standard Liquid filters are supported: upcase, downcase, date, strip_html, truncatewords, split, join, replace, and the rest of the standard set.

Example: Remap Fields

{
  "jobTitle": "{{ data.title }}",
  "employer": "{{ data.company.name }}",
  "location": "{{ data.locations[0].city }}, {{ data.locations[0].country }}",
  "applyUrl": "{{ data.apply_url }}",
  "postedAt": "{{ data.date_posted | date: '%Y-%m-%d' }}"
}

Example: Conditional Fields

{
  "title": "{{ data.title }}",
  "work_model": {% if data.is_remote %}"remote"{% else %}"{{ data.workplace_type | downcase }}"{% endif %},
  "skills": [{% for skill in data.all_skills %}"{{ skill }}"{% unless forloop.last %}, {% endunless %}{% endfor %}]
}

AI (OpenRouter)

Use LLM intelligence to extract, classify, enrich, or rewrite job data. The AI node sends data to any model available on OpenRouter — using your own OpenRouter API key, configured in Settings.

Configuration

Property	Type	Required	Description
`model`	string	Yes	OpenRouter model ID (e.g. `liquid/lfm2-8b-a1b`, `openai/gpt-4o-mini`, `anthropic/claude-3.5-sonnet`)
`system_prompt`	string	No	System instructions for the AI
`user_prompt`	string	Yes	The prompt template. Use `{{input}}` to inject the parent node’s output.
`output_schema`	object	No	JSON Schema enforced via OpenRouter structured outputs (strict mode). Strongly recommended.

The {{input}} placeholder in the user prompt is replaced with the full JSON output of the parent node. If the placeholder is missing, the input is appended to the end of the prompt automatically.

Runtime behavior

Structured output: with an output_schema, the request uses OpenRouter’s strict json_schema response format; without one, JSON mode is requested. If the model still returns plain text, it’s wrapped as { "result": "<text>" }.
Timeout & retries: 60-second timeout per attempt; transient failures (timeouts, network errors, 5xx, rate limits) are retried twice with backoff. Not configurable.
Key problems fail fast: an invalid key or exhausted credits (401/402/403) fails the entire sync immediately with an actionable message — it would otherwise fail identically on every job.

Example: Extract Seniority Signals

System Prompt:

You are a job data extraction assistant. Always respond with valid JSON.

User Prompt:

Extract the seniority level and required years of experience from this job posting:

{{input}}

Output Schema:

{
  "type": "object",
  "properties": {
    "seniority": { "type": "string", "enum": ["entry", "mid", "senior", "lead", "executive"] },
    "min_years_experience": { "type": ["integer", "null"] }
  },
  "required": ["seniority", "min_years_experience"],
  "additionalProperties": false
}

JSON Merge

Combine two JSON inputs into a single unified object. The JSON Merge node has two input handles: base (original data) and patch (new data, typically AI output). This is essential for preserving original job data while adding AI-generated fields.

Configuration

Property	Type	Default	Description
`strategy`	enum	`deep`	deep — recursively merges nested objects. shallow — only merges top-level keys.
`conflict_resolution`	enum	`patch_wins`	patch_wins — patch values override base on conflict. base_wins — base values are preserved on conflict.

Arrays are always replaced whole, never merged element-wise.

Deep Merge Example

Base (original data)
Patch (AI output)
Result (deep, patch wins)

{
  "title": "Software Engineer",
  "company": { "name": "Acme" }
}

{
  "skills": ["Python", "AWS"],
  "seniority": "senior"
}

{
  "title": "Software Engineer",
  "company": { "name": "Acme" },
  "skills": ["Python", "AWS"],
  "seniority": "senior"
}

Nested Deep Merge Example

When both base and patch have nested objects, deep merge recursively combines them:

Base
Patch
Result (deep, patch wins)

{
  "title": "SWE",
  "skills": { "hard": ["Python"] }
}

{
  "skills": { "hard": ["Python", "Go"], "soft": ["Leadership"] }
}

{
  "title": "SWE",
  "skills": { "hard": ["Python", "Go"], "soft": ["Leadership"] }
}

Use deep + patch wins (the defaults) in most cases. This preserves all original fields while letting AI-generated data override specific nested values.

Output Node

The exit point for processed data. The final JSON from this node is what gets written to your destination. Every pipeline ends with exactly one Output node — it has no configuration.

The output must be a JSON object (documents, not arrays or scalars), and its identity in the destination is always the original Jobo job id — regardless of what your pipeline renamed. That’s what makes updates and expiry deletions work no matter the output shape.

​Input Node

​The job document

​Filter

​Configuration

​Field paths

​Operators

​Example: remote jobs in the US or UK

​Field Mapper

​Configuration

​Behavior

​Example: flatten to a custom schema

​Liquid Template

​Available Variables

​Available Filters

​Example: Remap Fields

​Example: Conditional Fields

​AI (OpenRouter)

​Configuration

​Runtime behavior

​Example: Extract Seniority Signals

​JSON Merge

​Configuration

​Deep Merge Example

​Nested Deep Merge Example

​Output Node

Input Node

The job document

Filter

Configuration

Field paths

Operators

Example: remote jobs in the US or UK

Field Mapper

Configuration

Behavior

Example: flatten to a custom schema

Liquid Template

Available Variables

Available Filters

Example: Remap Fields

Example: Conditional Fields

AI (OpenRouter)

Configuration

Runtime behavior

Example: Extract Seniority Signals

JSON Merge

Configuration

Deep Merge Example

Nested Deep Merge Example

Output Node