Skip to main content
POST
Embed Resume
Upload a resume document in PDF, Word, text, or HTML format. The service parses structured data (personal info, experience, skills, etc.) and returns both the parsed resume and its vector embedding.
/embed is stateless and returns only an embedding (no stored id). If you want a reusable artifact_id for vsearch search_type: "resume" or neural search vector.artifact_id, use Upload Resume instead. Use /embed when you don’t want the resume stored server-side — pass result.embedding to POST /v2/jobs/neural-search as vector.vectors.

Endpoint

Request

Content-Type: multipart/form-data
Note:
  • The request must include exactly one file.
  • Supported file types: PDF, DOC/DOCX, plain text, HTML.
  • Maximum file size: 5 MB.
  • The uploaded file part must carry a correct Content-Type (e.g., application/pdf, text/plain, text/html, or the DOCX MIME type). Multipart requests that omit it (some HTTP clients default to application/octet-stream) are rejected with 400 File must be PDF or Word document, even for otherwise-valid files. cURL sets this automatically from the file extension.
file
required
The resume file to upload.

Response

object
Parsed resume data. All fields are optional and will appear only if detected in the document.
object
Embedding metadata and vector.

Example Request

Example Response

embedding contains 768 floats — only the first six are shown above for brevity.

Error Responses

Missing file:
  • Unsupported file type
  • File exceeds 5 MB limit
Unexpected error while parsing or embedding the resume.
Tip:
  • Ensure your resume is well-formatted to maximize accurate parsing.
  • All resume fields are optional; only detected sections will be returned.