README.md

metadata

title: Candidate Explorer
emoji: 🔍
colorFrom: purple
colorTo: gray
sdk: docker
app_port: 7860

Schemas

1. Build Schema
    a. table: cv_users ✅
        - user_id
        - username
        - hashed_password
        - email
        - full_name
        - role
        - is_active
        - tenant_id
        - created_at
        - date_modified
        - notes
    b. table: cv_tenant ✅
        - tenant_id
        - tenant_name
        - created_at
        - date_modified
        - notes
    c. table: cv_file ✅
        - file_id
        - file_type
        - filename
        - url
        - is_extracted
        - uploaded_at
        - date_modified
    d. table: cv_profile ✅
        - profile_id
        - fullname
        - gpa_edu_1
        - univ_edu_1
        - major_edu_1
        - gpa_edu_2
        - univ_edu_2
        - major_edu_2
        - gpa_edu_3
        - univ_edu_3
        - major_edu_3
        - domicile
        - yoe
        - hardskills
        - softskills
        - certifications
        - business_domain
        - filename
        - file_id
        - created_at
    e. table: cv_filter
        - criteria_id
        - gpa_edu_1
        - gpa_edu_2
        - gpa_edu_3
        - univ_edu_1
        - univ_edu_2
        - univ_edu_3
        - major_edu_1
        - major_edu_2
        - major_edu_3
        - domicile
        - yoe
        - hardskills
        - softskills
        - certifications
        - business_domain
        - created_at
    f. table: cv_weight ✅
        - weight_id
        - gpa_edu_1
        - gpa_edu_2
        - gpa_edu_3
        - univ_edu_1
        - univ_edu_2
        - univ_edu_3
        - major_edu_1
        - major_edu_2
        - major_edu_3
        - domicile
        - yoe
        - hardskills
        - softskills
        - certifications
        - business_domain
        - created_at
    g. table: cv_matching ✅
        - matching_id
        - profile_id
        - criteria_id
        - gpa_edu_1
        - gpa_edu_2
        - gpa_edu_3
        - univ_edu_1
        - univ_edu_2
        - univ_edu_3
        - major_edu_1
        - major_edu_2
        - major_edu_3
        - domicile
        - yoe
        - hardskills
        - softskills
        - certifications
        - business_domain
        - created_at
    g. table: profile_scores ✅
        - score_id
        - matching_id
        - score
        - created_at

List API

User

POST /admin/users              ✅  Create user
POST /admin/login              ✅  Login, returns JWT Bearer token
GET  /admin/me                 ✅  Get current authenticated user

Tenant

POST /admin/tenants            ✅  Create tenant
GET  /admin/tenants/{name}     ✅  Get tenant by name

File

POST   /file/upload            ✅  Upload multiple PDF files
GET    /file/user/{user_id}    ✅  Get file metadata by user
DELETE /file/{filename}        ✅  Soft-delete file (marks is_deleted=True)
GET    /file/score_card        ✅  Dashboard summary (total files, extracted profiles, %)

Profile

POST /profile/extract_profile          ✅  Extract structured profile from a CV PDF
GET  /profile/{profile_id}             ✅  Get extracted profile by ID
GET  /profile/profiles/{criteria_id}   ✅  Get profiles by criteria ID

Agentic

POST /agentic/create_filter            ✅          Create filter/criteria
GET  /agentic/filter/{criteria_id}     ✅          Get filter by ID
POST /agentic/create_weight            ✅          Create weight config for scoring
POST /agentic/calculate_score          ⚠️ DEPRECATED  Old single scoring endpoint
POST /agentic/v2/calculate_score       ✅          Bulk scoring (v2, recommended)

How To

1. Init schema on DB uv run python -m externals.databases.pg_schema_up

2. Drop schema on DB uv run python -m externals.databases.pg_schema_down

Compile pyproject.toml to requirements.txt uv pip compile pyproject.toml -o requirements.txt