Ceevee career intel

CeeVee Career Intel — AI Career Intelligence Chat

Start a career intelligence conversation with live web research, knowledge base context, and optional CV-awareness for personalized insights.

Workflow

Step 1: Set Up Context (FREE)

Check for existing CV (optional but recommended):

1. Call ceevee_list_versions to see if the user has uploaded CVs
2. If a CV exists, save the cv_version_id of the main version
   • CV context makes responses personalized to the user's seniority, industry, and skills
3. If no CV exists, that's fine — career intel works without one

Check for existing conversations:

1. Optionally call ceevee_list_versions to check for prior chat history
   • The user may want to continue an existing conversation thread

Step 2: Start the Conversation (2 credits/message)

If the user provided a topic as $ARGUMENTS, use it as the first message. Otherwise, ask what they want to explore.

Suggested starter topics:

• "What's the salary range for [role] in [location]?"
• "What skills are most in-demand for [industry] in 2026?"
• "How is AI disrupting [function/industry]?"
• "What career paths lead from [current role] to [target role]?"
• "What should I negotiate for in a [role] offer?"
• "Compare compensation packages: [Company A] vs [Company B]"
• "What certifications matter most for [field]?"
• "How does the job market look for [role] in the next 12 months?"

Call ceevee_chat with:

• message — The user's question or topic
• cv_version_id — If available from Step 1
• conversation_id — Omit for first message, include for follow-ups

The response includes:

• AI analysis grounded in live web search and knowledge base
• Sources — Citations from both KB documents and live search results
• Conversation ID — Save this for follow-up messages

Step 3: Continue the Thread (2 credits/message)

For follow-up questions, always pass the conversation_id from Step 2 to maintain conversation context.

The AI retains context from prior messages in the thread, so the user can:

• Drill deeper into a topic ("Tell me more about the AI skills gap")
• Compare alternatives ("How does that compare to cloud architecture roles?")
• Get actionable advice ("What steps should I take in the next 3 months?")
• Request specifics ("What companies in London are hiring for this?")

Step 4: Alternative Channels

Depending on what the user needs, other tools may be more appropriate:

For compensation-specific advice:

• Use atlas_advisor_chat (2 credits) for lightweight hiring/compensation questions
• Use atlas_chat (3 credits) for workforce intelligence with live tool access to salary benchmarking, supply/demand analysis, and competitor intel

For CV-specific optimization:

• Direct to /ceevee-optimize for the full positioning pipeline

For comprehensive research:

• Direct to /atlas-report for a full PDF research report (15 credits)

Credit Cost

Action	Cost
Check CV versions	FREE
Each chat message	2 credits
Explain a CV edit	1 credit

Typical session: 3-5 messages = 6-10 credits

What Career Intel Covers

The AI combines three intelligence sources for every response:

1. Knowledge Base — Curated career intelligence from HBR, McKinsey, BCG, Gartner, and 50+ premium sources, refreshed 3x daily
2. Live Web Search — Real-time Tavily search for current salary data, job postings, and market trends
3. CV Context (when available) — Personalized to the user's experience level, industry, function, and skill profile

Best for:

• Salary and compensation intelligence
• Market trends and industry disruption analysis
• Skill gap identification and upskilling priorities
• Career trajectory planning and pivot analysis
• Negotiation preparation and offer comparison
• Industry-specific hiring trends

Not for:

• CV editing (use /ceevee-optimize)
• Candidate evaluation (use /atlas-gem, /atlas-fit, or /atlas-shortlist)
• Full research reports (use /atlas-report)

Tips

• Including a CV makes responses significantly more relevant — the AI calibrates to seniority and industry
• Career intel automatically stores live search findings to the knowledge base, improving future responses
• For salary questions, be specific about role, location, and seniority for the most accurate data
• Conversations are persistent — you can return to a thread later by passing the same conversation_id
• Each message costs 2 credits regardless of response length — batch related questions together