LinkedIn Multi-Mode Actor (raw‑expanded) 🚀

Pull profiles, companies, and posts from LinkedIn through a single actor.
Outputs include the original raw fields expanded at top level with a data_ prefix (e.g., data_followers, data_company_size).
This makes it easy to query/filter without digging into nested blobs.

---

✨ Highlights

• One input, many modes → profiles, companies, a single post, or recent posts by profile.
• Raw-expanded output → every key from the upstream raw payload is duplicated at top level with the data_ prefix.
• Cache-aware → repeated identical requests can return instantly with status: "cached".
• Safety clamp → limit ≤ 50 enforced on multi-result modes.
• Tenant scoping (optional) → pass a userId to partition cache per tenant/workspace.
• Operational guidance → Good overall timeout budget is ~20 minutes for reliable completion on fresh, uncached targets.

---

🧭 Supported modes

mode	purpose	required fields
profileByUrl	Fetch a profile by URL	url
profileByName	Search profiles by full name	fullName or fullNames[], limit?
companyByUrl	Fetch a company by URL	url
postByUrl	Fetch a single public post/article	url
postsByProfile	Recent posts for a given profile URL	profileUrl or profileUrls[], limit?

---

⚙️ Inputs

Provide a single JSON object. Only fill fields relevant to your chosen mode.

Common

• mode (string, required) — one of the modes above.
• limit (integer, optional, 1–50) — cap for multi-result modes (profileByName, postsByProfile). Default 10.

Batching

• urls (array) — for profileByUrl, companyByUrl, postByUrl.
• fullNames (array) — for profileByName.
• profileUrls (array) — for postsByProfile.

Single-value

• url (string) — entity URL for profileByUrl, companyByUrl, postByUrl.
• fullName (string) — for profileByName.
• profileUrl (string) — for postsByProfile.

Tip: Prefer canonical URLs on https://www.linkedin.com/... (avoid country subdomains and tracking params).

---

🏃 Quick start

Profile by URL

{ "mode": "profileByUrl", "url": "https://www.linkedin.com/in/reidhoffman/" }

Company by URL

{ "mode": "companyByUrl", "url": "https://www.linkedin.com/company/ibm/" }

Search by name

{ "mode": "profileByName", "fullName": "Ada Lovelace", "limit": 3 }

Recent posts for a profile

{ "mode": "postsByProfile", "profileUrl": "https://www.linkedin.com/in/satyanadella/", "limit": 10 }

Batch

{
  "mode": "postsByProfile",
  "profileUrls": [
    "https://www.linkedin.com/in/reidhoffman/",
    "https://www.linkedin.com/in/sundarpichai/"
  ],
  "limit": 8
}

---

📦 Output examples (raw‑expanded)

Below are representative examples. Fields coming from the upstream raw payload appear with the data_ prefix.

Profile (single-entity)

{
  "jobId": "68cfc1562c409e61554902e1",
  "mode": "profileByUrl",
  "status": "succeeded",
  "type": "entity",
  "key": "li:profile:https://www.linkedin.com/in/ada-lovelace",
  "canonicalUrl": "https://www.linkedin.com/in/ada-lovelace",
  "fetchedAt": "2025-09-21T09:20:00.000Z",

  "fullName": "Ada Lovelace",
  "headline": "Pioneer of Computing",
  "location": "London, United Kingdom",
  "experiences": [
    { "title": "Chief Visionary", "company": "Analytical Engines Ltd", "startDate": "1842-01", "endDate": null, "location": "London" }
  ],
  "educations": [
    { "school": "Self-taught", "degree": null, "field": "Mathematics", "startDate": null, "endDate": null }
  ],

  "data_full_name": "Ada Lovelace",
  "data_headline": "Pioneer of Computing",
  "data_location": "London, United Kingdom",
  "data_experiences": [
    { "title": "Chief Visionary", "company": "Analytical Engines Ltd", "start_date": "1842-01" }
  ],
  "data_education": [
    { "school": "Self-taught", "field": "Mathematics" }
  ],
  "data_url": "https://www.linkedin.com/in/ada-lovelace"
}

Posts by profile (multi-result — one dataset item per post)

{
  "jobId": "68cfc16d8e1ddf4d9a4e3e2a",
  "mode": "postsByProfile",
  "status": "succeeded",
  "type": "item",
  "key": "li:post:https://www.linkedin.com/posts/1234567890",
  "canonicalUrl": "https://www.linkedin.com/posts/1234567890",
  "fetchedAt": "2025-09-21T09:20:03.000Z",

  "text": "Reflections on analytical computation…",
  "engagement": { "likes": 42, "comments": 7, "reposts": 3 },

  "data_post_id": "1234567890",
  "data_text": "Reflections on analytical computation…",
  "data_likes_count": 42,
  "data_comments_count": 7,
  "data_reposts_count": 3,
  "data_post_url": "https://www.linkedin.com/posts/1234567890",
  "data_images": [],
  "data_videos": []
}

Company (single-entity)

{
  "jobId": "68cfc17c1c9c2f0b6d7c8f21",
  "mode": "companyByUrl",
  "status": "succeeded",
  "type": "entity",
  "key": "li:company:https://www.linkedin.com/company/openai",
  "canonicalUrl": "https://www.linkedin.com/company/openai",
  "fetchedAt": "2025-09-21T09:20:10.000Z",

  "name": "OpenAI",
  "website": "https://openai.com",
  "industries": "Artificial Intelligence",
  "companySize": "1001-5000 employees",

  "data_name": "OpenAI",
  "data_website": "https://openai.com",
  "data_followers": 2500000,
  "data_company_size": "1001-5000 employees",
  "data_logo": "https://media.licdn.com/...",
  "data_image": "https://media.licdn.com/...",
  "data_about": "Creating safe AGI that benefits all of humanity",
  "data_country_code": "US",
  "data_url": "https://www.linkedin.com/company/openai"
}

Cached response

{
  "jobId": "68cfc188aef2b963e38e1f40",
  "mode": "profileByUrl",
  "status": "cached",
  "type": "entity",
  "key": "li:profile:https://www.linkedin.com/in/ada-lovelace",
  "cache": { "hit": true }
}

On timeouts or empty terminal states the actor pushes a clear record (e.g., { status: "timeout" }) so runs don’t crash.

---

🟡 Status values

status	meaning
succeeded	Full data collected
partial	Some data available (e.g., early items before finish)
failed	Unrecoverable error
cached	Served from recent identical request

---

🔧 Environment knobs (optional)

• API_KEY / API_KEYS — backend API key (first of API_KEYS is used).
• SERVER_BASE — override the default server URL.
• VERBOSE — set to 0/false to silence periodic poll logs.
• POLL_INTERVAL_MS — poll cadence (default 5000 ms).
• TIMEOUT_MS — per-job overall wait. Recommended budget ~1,200,000 ms (20 minutes) for fresh, uncached targets in busy hours.

The actor handles timeouts gracefully: it logs and emits a dataset row instead of throwing.

---

🧪 More input snippets

Profiles by name (batch)

{ "mode": "profileByName", "fullNames": ["Diego Fernandez", "Anya Petrova", "Hassan Ali"], "limit": 3 }

Posts by profile (batch)

{
  "mode": "postsByProfile",
  "profileUrls": [
    "https://www.linkedin.com/in/dharmesh/",
    "https://www.linkedin.com/in/satyanadella/"
  ],
  "limit": 8
}

---

🛡️ Compliance

Please comply with LinkedIn’s Terms of Service and applicable laws.
This actor does not bypass access controls.

---

🧰 Support

Open an issue in the Issues tab. Include your input JSON and a redacted log excerpt for fastest help.