Work & Workforce Engineering · Agent

Skill Proficiency Analyzer

A chat agent that reads professional documents, identifies every skill, and rates each one Expert, Advanced, Proficient, or Beginner - one file at a time or in bulk.

Documentation Pricing

Type Conversational agent Status Stable Pricing unit Per MB of input Access Workspace

What it does

Share a resume or professional document and the agent extracts every skill, weighs the evidence behind it, and rates each one Expert, Advanced, Proficient, or Beginner. You get a structured proficiency profile you can then explore in plain language - ask about a specific skill, a level, the strongest areas.

Analyze a single document on the fly, or bulk-upload a whole batch - several files or a ZIP - where each document is analysed as its own profile and returned in one combined, downloadable result. Ratings are grounded in evidence from the document itself, not self-reported levels.

Analysis modes

Switch between two modes any time from the toggle at the top of the chat:

Guided Analysis. A structured, step-by-step flow: upload a document (or ZIP) and get a consistent proficiency breakdown across Expert, Advanced, Proficient, and Beginner. Follow-up questions about the result are free. Best for quick, repeatable reports.
Ask Anything. An open, AI-led conversation over the profiles you've uploaded in the chat. Ask anything in your own words - by text or voice - and it looks up your profiles and runs analytics across them: compare profiles, find common or missing skills, and spot gaps, replying in your language. Best for exploring and analysing.

Capabilities

Skill extraction & rating. Detects explicit and implied skills and maps each to a proficiency level (Expert · Advanced · Proficient · Beginner) from contextual evidence.
Bulk analysis. Upload multiple files or a ZIP - each document is analysed individually as a unique profile, with a combined result you can download together.
Conversational follow-up. Ask about any skill, level, the strongest/weakest areas, a summary, or a score - all in chat.
Export. Download a result as PDF, JSON, or CSV; export a whole batch as one combined CSV or a ZIP.

Inputs & outputs

Inputs

Name	Type	Required	Description
document	PDF · DOC · DOCX · ZIP	required	The document(s) to analyse. A ZIP is unpacked and each file inside is analysed as its own profile.
message	string	optional	A question or instruction to steer the conversation (e.g. "show the strongest skills").

Outputs

Name	Type	Description
bands	array<object>	Skills grouped into Expert (L4), Advanced (L3), Proficient (L2), and Beginner (L1) levels.
skillCount	integer	Total number of skills identified in the document.
summary	string	A plain-language read-out of the profile you can follow up on in chat.

Example tasks

Things you can hand this agent:

"Analyse this resume and show me the Expert-level skills."
"Upload this ZIP of 10 candidates and tell me who has the strongest cloud skills."
"Which skills in this profile need the most improvement, and give it a score out of 10."

Example result (illustrative)

{
  "fileName": "resume.pdf",
  "skillCount": 13,
  "bands": [
    { "level": "L4", "label": "Expert",     "skills": ["Testing", "Postman", "Agile"] },
    { "level": "L3", "label": "Advanced",   "skills": ["Android", "SQL", "Scrum"] },
    { "level": "L2", "label": "Proficient", "skills": ["Automation", "Reporting"] },
    { "level": "L1", "label": "Beginner",   "skills": ["HTTP Request", "E-commerce"] }
  ]
}

Configuration

The agent is conversational and runs with no setup - these are the formats it accepts and produces:

Option	Type	Default	Description
formats	array<string>	pdf, doc, docx, zip	Accepted document types. A ZIP may contain only PDF, DOC, or DOCX files.
export	array<string>	pdf, json, csv	Download formats for a result. A bulk run can also export one combined CSV or a ZIP.

No setup needed. Just upload a document in chat and ask follow-up questions. Results are downloadable and saved to your history.

Charges

Two independent charges. The document-upload charge is the same in both modes; the only extra in Ask Anything is a small per-question conversation charge.

What	Applies to	Charge
Document upload	Both modes (Guided Analysis & Ask Anything)	25 credits ($2.50) per MB of input - a document up to 1 MB counts as 1 MB, then per additional MB. ZIP uploads are billed on the uncompressed size, and only successfully processed files are charged.
Conversation message	Ask Anything only	2 credits ($0.13) per question you ask in the conversation.
Follow-up question	Guided Analysis	Free - follow-up questions about a Guided result are not charged.

Example. A 1 MB resume analysed in either mode costs 25 credits ($2.50). In Ask Anything, each question you then ask about it adds 5 credits ($0.50); in Guided Analysis those follow-ups are free. Exact rates are shown live in the agent and on the pricing page.

Limits & guardrails

Supported inputs are PDF, DOC, DOCX, or a ZIP containing those file types.
In a ZIP, each document is analysed as its own unique profile - never a single merged one.
Only successfully processed files are charged - failed files are not charged.
Documents are processed only to return your result and are never used to train AI models.