humanmint

HumanMint v2

HumanMint cleans and normalizes messy contact data with one call. It standardizes names, emails, phones, addresses, departments, titles, and organizations. It's built for both public-sector data and B2B (CEOs, VPs, directors, managers) and ships with curated public-sector mappings.

from humanmint import mint

result = mint(
    name="Dr. John Q. Smith, PhD",
    email="[e-mail]",
    phone="[phone] ext 456",
    department="001 - Public Works Dept",
    title="Chief of Police",
    address="123 N. Main St Apt 4B, Madison, WI 53703",
    organization="City of Madison Police Department",
)

result.name_standardized          # "John Q Smith"
result.email_standardized         # "[e-mail]"
result.phone_pretty               # "[phone]"
result.department_canonical       # "Public Works"
result.title_canonical            # "police chief"
result.address_canonical          # "123 N. Main St Apt 4B Madison WI 53703 US"

Multi-person splitting:

mint(name="John and Jane Smith", split_multi=True)
# -> [MintResult(John Smith), MintResult(Jane Smith)]

Why HumanMint

• General-purpose: works for government and B2B without swapping libraries.
• Real-world chaos: handles titles inside names, departments with codes/phones, smashed addresses, anti-scraper emails, casing quirks.
• Unique data: 23K+ department variants -> 64 categories; 73K+ titles with curated canonicals + BLS; context-aware title mapping.
• Safe defaults: length guards, optional aggressive cleaning, semantic conflict checks, bulk dedupe, multi-person name splitting.
• Fast: lazy imports for quick startup, process-based bulk for CPU-bound speed, built-in dedupe to avoid redundant work.

AI extraction (optional)

Install the ML extra (pip install humanmint[ml]) and pass text= with use_gliner=True to extract from unstructured text, then normalize. Structured fields you pass always win. GLiNER extraction is experimental; prefer structured inputs when available.

from humanmint.gliner import GlinerConfig
result = mint(text=signature_block, use_gliner=True, gliner_cfg=GlinerConfig(threshold=0.85))

Installation

pip install humanmint
# Optional extras:
#   pip install humanmint[address]  # usaddress parsing
#   pip install humanmint[pandas]   # DataFrame helpers
#   pip install humanmint[ml]       # GLiNER2 extraction

Quickstart

from humanmint import mint, compare, bulk

r1 = mint(name="Jane Doe", email="[e-mail]", department="Public Works", title="Engineer")
r2 = mint(name="J. Doe",  email="[e-mail]", department="PW Dept",       title="Public Works Engineer")

score, why = compare(r1, r2, explain=True)

records = [
    {"name": "Alice", "email": "[e-mail]"},
    {"name": "Bob",   "email": "[e-mail]"},
]
results = bulk(records, workers=4)

Access Patterns

• Dicts: result.title["canonical"], result.department["canonical"], result.department["category"]
• Properties: name_standardized, title_canonical, department_canonical, email_standardized, phone_standardized, address_canonical, organization_canonical
• Full dicts: result.title, result.department, result.email, etc.

Recommended Properties

• Names: name_standardized, name_first, name_last, name_middle, name_suffix, name_gender, name_nickname
• Name extras: name_salutation (Mr./Ms./Mx.)
• Emails: email_standardized, email_domain, email_is_valid, email_is_generic_inbox, email_is_free_provider
• Phones: phone_standardized, phone_e164, phone_pretty, phone_extension, phone_is_valid, phone_type, phone_location, phone_time_zones
• Departments: department_canonical, department_category, department_normalized, department_override
• Titles: title_canonical, title_raw, title_normalized, title_is_valid, title_confidence, title_seniority
• Addresses: address_canonical, address_raw, address_street, address_unit, address_city, address_state, address_zip, address_country
• Organizations: organization_raw, organization_normalized, organization_canonical, organization_confidence

Use result.get("email.is_valid") to fetch nested dict values via dot paths.

Comparing Records

from humanmint import compare
score, reasons = compare(r1, r2, explain=True)  # 0->100

Batch & Export

from humanmint import bulk, export_json, export_csv, export_parquet, export_sql

# Process records in parallel
results = bulk(records, workers=4, progress=True)

# Export results to various formats
export_json(results, "out.json")
export_csv(results, "out.csv", flatten=True)

# Note: For per-record overrides (dept_overrides, title_overrides), include them in each record dict
records_with_overrides = [
    {**rec, "dept_overrides": {"IT": "Information Technology"}}
    for rec in records
]
results = bulk(records_with_overrides, workers=4)

CLI

humanmint clean input.csv output.csv --name-col name --email-col email --phone-col phone --dept-col department --title-col title

Performance (current)

• Cold import: ~0.5 s (with pandas installed).
• First call warm-up: ~0.5 s (loads caches).
• Bulk: process-based parallelism; throughput scales with cores and workload size.

Notes

• US-focused address parsing; usaddress used when available, otherwise heuristics.
• Optional deps (pandas, pyarrow, sqlalchemy, rich, tqdm) enhance exports and progress bars.
• Department and title datasets are curated and updated regularly for best accuracy.