🕵️ Rewriting Biographies¶

Instead of replacing entities with tokens, rewrite mode generates a privacy-safe transformation of the entire text. The run() / preview() pipeline:

1. Detects entities (same as replace mode, plus latent entity detection)
2. Classifies the domain and assigns sensitivity dispositions
3. Generates a rewritten version that obscures sensitive entities
4. Evaluates quality (utility) and privacy (leakage) with an automated repair loop

Afterward, a separate optional evaluate() call runs LLM judges for detection validity and holistic privacy, quality, and style scores.

📚 What you'll learn¶

• Configure rewrite mode with PrivacyGoal to specify what to protect and what to preserve
• Set evaluation criteria and risk tolerance for automated quality checks
• Preview rewritten text and inspect utility / leakage scores
• Triage flagged records with needs_human_review
• Run evaluate() for detection validity and holistic judge scores (privacy, quality, style)

Tip: First time running notebooks? Start with setup instructions.

⚙️ Setup¶

• Check if your NVIDIA_API_KEY from build.nvidia.com is registered for model access.
  • The default build.nvidia.com (NVIDIA Build) setup is a convenient way to try Anonymizer and iterate on previews. Use of NVIDIA Build is subject to NVIDIA Build's own terms of service and privacy practices, which are separate from and independent of the NeMo Framework library. NVIDIA Build is intended for evaluation and testing purposes only and may not be used in production environments. Do not upload any confidential information or personal data when using NVIDIA Build. Your use of NVIDIA Build is logged for security purposes and to improve NVIDIA products and services.
  • Request and token rate limits on build.nvidia.com vary by account and model access, and lower-volume development access can be slow for full-dataset runs. Start with preview() on a small sample, then move to your own endpoint for production data and usage.
• Import Rewrite and PrivacyGoal.
• Anonymizer() initializes with the default model provider -- no extra config needed.
• configure_logging(LoggingConfig.default()) keeps logs at INFO. Switch to LoggingConfig.debug() when troubleshooting.

In [2]:

import getpass
import os

if not os.getenv("NVIDIA_API_KEY"):
    key = getpass.getpass("Enter NVIDIA_API_KEY from build.nvidia.com: ").strip()
    if not key:
        raise RuntimeError("NVIDIA_API_KEY is required to run these notebooks.")
    os.environ["NVIDIA_API_KEY"] = key

import getpass import os if not os.getenv("NVIDIA_API_KEY"): key = getpass.getpass("Enter NVIDIA_API_KEY from build.nvidia.com: ").strip() if not key: raise RuntimeError("NVIDIA_API_KEY is required to run these notebooks.") os.environ["NVIDIA_API_KEY"] = key

In [3]:

from anonymizer import (
    Anonymizer,
    AnonymizerConfig,
    AnonymizerInput,
    LoggingConfig,
    PrivacyGoal,
    Rewrite,
    configure_logging,
)

configure_logging(LoggingConfig.default())

from anonymizer import ( Anonymizer, AnonymizerConfig, AnonymizerInput, LoggingConfig, PrivacyGoal, Rewrite, configure_logging, ) configure_logging(LoggingConfig.default())

In [4]:

anonymizer = Anonymizer()

anonymizer = Anonymizer()

[11:14:00] [INFO] 🔧 Anonymizer initialized with 3 model configs

[11:14:00] [INFO]   |-- 🔎 detector:  gliner-pii-detector

[11:14:00] [INFO]   |-- ✅ validator: gpt-oss-120b

[11:14:00] [INFO]   |-- 🧩 augmenter: gpt-oss-120b

📦 Input data¶

• Same biographies dataset used in earlier notebooks -- familiar data makes it easy to compare rewrite output against replace output.

In [5]:

input_data = AnonymizerInput(
    source="https://raw.githubusercontent.com/NVIDIA-NeMo/Anonymizer/refs/heads/main/docs/data/NVIDIA_synthetic_biographies.csv",
    text_column="biography",
    data_summary="Biographical profiles",
)

input_data = AnonymizerInput( source="https://raw.githubusercontent.com/NVIDIA-NeMo/Anonymizer/refs/heads/main/docs/data/NVIDIA_synthetic_biographies.csv", text_column="biography", data_summary="Biographical profiles", )

🎛️ Configure¶

• PrivacyGoal spells out what to protect and what to preserve -- this gives the rewriter clear, domain-specific guidance.
• risk_tolerance (default "low") and max_repair_iterations (default 3) control the automated quality gate -- see Risk tolerance for presets.

In [6]:

config = AnonymizerConfig(
    rewrite=Rewrite(
        privacy_goal=PrivacyGoal(
            protect="All direct identifiers and quasi-identifier combinations (names, locations, employers, dates)",
            preserve="Career trajectory, educational background, and professional accomplishments",
        ),
        risk_tolerance="low",
        max_repair_iterations=3,
    ),
)

config = AnonymizerConfig( rewrite=Rewrite( privacy_goal=PrivacyGoal( protect="All direct identifiers and quasi-identifier combinations (names, locations, employers, dates)", preserve="Career trajectory, educational background, and professional accomplishments", ), risk_tolerance="low", max_repair_iterations=3, ), )

👁️ Preview¶

• preview() runs on a small sample so you can iterate on privacy goals and evaluation criteria before committing to a full run.

In [7]:

preview = anonymizer.preview(
    config=config,
    data=input_data,
    num_records=3,
)

preview.display_record(0)

preview = anonymizer.preview( config=config, data=input_data, num_records=3, ) preview.display_record(0)

[11:14:00] [INFO] 👀 Preview mode: 📂 Loaded 3 records from https://raw.githubusercontent.com/NVIDIA-NeMo/Anonymizer/refs/heads/main/docs/data/NVIDIA_synthetic_biographies.csv (column: 'biography')

[11:14:00] [INFO] 🔍 Running entity detection on 3 records

[11:14:00] [INFO] detection labels in scope: (default: 65 labels; see anonymizer.DEFAULT_ENTITY_LABELS for list)

[11:15:03] [INFO]   |-- 📋 Detection complete — 79 entities found across 3 records (0 failed) [62.6s]

[11:15:03] [INFO]   |-- labels: first_name=23, organization_name=8, age=5, occupation=5, city=4, state=4, degree=4, university=4, field_of_study=4, last_name=3, race_ethnicity=3, political_view=3, language=2, religious_belief=2, street_address=2, place_name=1, date_of_birth=1, employment_status=1

[11:15:03] [INFO] ✏️ Running rewrite pipeline

[11:19:15] [INFO] Evaluate-repair loop: all rows pass at iteration 0

[11:19:15] [INFO]   |-- 📋 Rewrite complete (0 failed) [251.7s]

[11:19:15] [INFO] 🎉 Pipeline complete — 3 records processed, 0 total failures

Anonymizer Rewrite Preview (record 0)

Original

Bobby| first_name Watford| last_name, a 40| age‑year‑old Mexican| race_ethnicity veterinarian| occupation living in Denver| city, Colorado| state, grew up on the outskirts of the city and developed a love for animals early on. After graduating from Jefferson High| organization_name, he earned his DVM| degree at the University of Colorado Boulder| university, where he also completed a research stint in wildlife health| field_of_study. Fluent in English| language, Bobby| first_name has always described his upbringing as a blend of small‑town curiosity and the vibrant culture of his community, values that continue to shape his compassionate approach to animal care. Since finishing his training, Bobby| first_name has worked at VCA Animal Hospital| organization_name and later at the Colorado Veterinary Clinic| organization_name, where he now leads a busy mixed‑practice team. He identifies as a Christian Democrat| political_view and often volunteers at local shelters, a habit encouraged by his wife, Maya| first_name, and their two teenage children, Aria| first_name and Leo| first_name. Outside the clinic, Bobby| first_name enjoys hiking the Rockies| place_name with his family and mentoring veterinary students from his alma mater.

Rewritten

Ethan Hawthorne, a veterinarian in his 40s of Mexican heritage living in a city in Colorado, grew up on the outskirts of the city and developed a love for animals early on. After graduating from Jefferson High, he earned his DVM at a Colorado university, where he also completed a research stint in wildlife health. Fluent in English, Ethan has always described his upbringing as a blend of small‑town curiosity and the vibrant culture of his community, values that continue to shape his compassionate approach to animal care. Since finishing his training, Ethan has worked at a regional veterinary hospital chain and later at a local veterinary clinic, where he now leads a busy mixed‑practice team. He identifies as a Christian Democrat and often volunteers at local shelters, a habit encouraged by his partner, Nina, and their two teenage children, Sofia and Jasper. Outside the clinic, Ethan enjoys hiking the Rockies with his family and mentoring veterinary students from his alma mater.

Scores

Utility: 1.00Leakage: 0.00Weighted Leakage Rate: 0.00Rewrite Needs Review: No

Entity Disposition

Entity	Label	Sensitivity	Protection
Bobby	first_name	high	replace
Watford	last_name	high	replace
40	age	high	generalize
Mexican	race_ethnicity	medium	leave_as_is
veterinarian	occupation	low	leave_as_is
Denver	city	high	generalize
Colorado	state	low	leave_as_is
Jefferson High	organization_name	low	leave_as_is
DVM	degree	medium	leave_as_is
University of Colorado Boulder	university	high	generalize
wildlife health	field_of_study	low	leave_as_is
English	language	low	leave_as_is
Christian Democrat	political_view	low	leave_as_is
Rockies	place_name	low	leave_as_is
VCA Animal Hospital	organization_name	high	generalize
Colorado Veterinary Clinic	organization_name	high	generalize
Maya	first_name	high	replace
Aria	first_name	high	replace
Leo	first_name	high	replace
large veterinary hospital chain	employer	medium	leave_as_is
married	marital_status	medium	suppress_inference
doctorate degree	education_level	low	leave_as_is
male	gender	low	leave_as_is

In [8]:

preview.display_record(1)

preview.display_record(1)

Anonymizer Rewrite Preview (record 1)

Original

Idilio| first_name Bell| last_name is a 37| age‑year‑old astronomer| occupation living in Edison| city, New Jersey| state. Born on November 21, 1988| date_of_birth, he grew up in a bilingual Italian| race_ethnicity household and speaks English| language at home and work. He earned his bachelor’s degree| degree in physics| field_of_study from the University of New Jersey| university and later completed a PhD| degree in astrophysics| field_of_study at Princeton| university, where his dissertation focused on exoplanet atmospheres. After graduation he spent three years at NASA’s Goddard Space Flight Center| organization_name before joining SpaceX| organization_name’s research division, where he now leads a team analyzing data from the Starlink telescope array| organization_name. Idilio| first_name describes himself as secular| religious_belief and leans progressive| political_view on most political issues, often volunteering for science outreach programs in his community. Outside the lab, Idilio| first_name shares a modest house on West Roberts Drive| street_address with his wife, Maya| first_name, and their two young daughters, Lina| first_name and Zara| first_name. His mother, Elena| first_name, lives nearby and still cooks the family’s favorite pasta on Sundays, while his father, Marco| first_name, retired| employment_status from an engineering firm in New York| state. Family gatherings are a mix of lively conversation and stargazing sessions on the backyard deck, where Idilio| first_name points out constellations and tells stories of the cosmos that inspire his children’s curiosity.

Rewritten

Dario Kumar is an astronomer in his late 30s living in a city in New Jersey. He grew up in a bilingual Italian household and speaks English at home and work. He earned his bachelor's degree in physics from the University of New Jersey and later completed a PhD in astrophysics at Princeton, where his dissertation focused on exoplanet atmospheres. After graduation he spent three years at NASA’s Goddard Space Flight Center before joining SpaceX’s research division, where he now leads a team analyzing data from the Starlink telescope array. Dario describes himself as secular and leans progressive on most political issues, often volunteering for science outreach programs in his community. Outside the lab, Dario shares a modest house on Maple Hill Lane with his wife, Aisha, and their two young daughters, Nina and Mila. His mother, Sofia, lives nearby and still cooks the family’s favorite pasta on Sundays, while his father, Luis, retired from an engineering firm in New York. Family gatherings are a mix of lively conversation and stargazing sessions on the backyard deck, where Dario points out constellations and tells stories of the cosmos that inspire his children’s curiosity.

Scores

Utility: 0.96Leakage: 0.00Weighted Leakage Rate: 0.00Rewrite Needs Review: No

Entity Disposition

Entity	Label	Sensitivity	Protection
Idilio	first_name	high	replace
Bell	last_name	high	replace
37	age	high	generalize
Edison	city	high	generalize
Elena	first_name	high	replace
English	language	low	leave_as_is
Italian	race_ethnicity	medium	leave_as_is
Lina	first_name	high	replace
Marco	first_name	high	replace
Maya	first_name	high	replace
NASA’s Goddard Space Flight Center	organization_name	high	leave_as_is
New Jersey	state	medium	leave_as_is
New York	state	medium	leave_as_is
November 21, 1988	date_of_birth	high	replace
PhD	degree	medium	leave_as_is
Princeton	university	high	leave_as_is
SpaceX	organization_name	high	leave_as_is
Starlink telescope array	organization_name	high	leave_as_is
University of New Jersey	university	high	leave_as_is
West Roberts Drive	street_address	high	replace
Zara	first_name	high	replace
astronomer	occupation	medium	leave_as_is
bachelor’s degree	degree	medium	leave_as_is
in astrophysics	field_of_study	medium	leave_as_is
in physics	field_of_study	medium	leave_as_is
progressive	political_view	low	leave_as_is
retired	employment_status	low	leave_as_is
secular	religious_belief	low	leave_as_is
married	marital_status	medium	leave_as_is
2	number_of_children	medium	leave_as_is
male	gender	low	leave_as_is
aerospace	industry	high	leave_as_is
doctorate	highest_education	high	leave_as_is
research team lead	current_role	medium	leave_as_is
Italian-American	cultural_background	medium	leave_as_is
single-family home	housing_type	low	leave_as_is

How to interpret leakage: Leakage is measured against the sensitivity disposition. Details marked leave_as_is may remain without increasing leakage_mass. If an output retains something you expected the privacy goal to protect, inspect the Entity Disposition table.

🚀 Full run¶

• result.dataframe has user-facing columns: rewritten text, scores, and the review flag.
• result.trace_dataframe has every intermediate column for debugging.

In [9]:

result = anonymizer.run(config=config, data=input_data)

result.dataframe.head()

result = anonymizer.run(config=config, data=input_data) result.dataframe.head()

[11:19:15] [INFO] 📂 Loaded 25 records from https://raw.githubusercontent.com/NVIDIA-NeMo/Anonymizer/refs/heads/main/docs/data/NVIDIA_synthetic_biographies.csv (column: 'biography')

[11:19:15] [INFO] 🔍 Running entity detection on 25 records

[11:19:15] [INFO] detection labels in scope: (default: 65 labels; see anonymizer.DEFAULT_ENTITY_LABELS for list)

[11:22:42] [INFO]   |-- 📋 Detection complete — 672 entities found across 25 records (0 failed) [207.0s]

[11:22:42] [INFO]   |-- labels: first_name=154, organization_name=63, occupation=47, city=41, field_of_study=38, university=36, race_ethnicity=30, last_name=27, age=27, degree=27, state=25, political_view=25, religious_belief=24, street_address=23, language=19, place_name=17, employment_status=11, county=10, date_of_birth=9, company_name=5, date=5, education_level=4, time=1, landmark=1, country=1, gender=1, postcode=1

[11:22:42] [INFO] ✏️ Running rewrite pipeline

[11:34:02] [INFO] Evaluate-repair loop iteration 0: 4/25 rows need repair

[11:34:44] [INFO] Evaluate-repair loop: all rows pass at iteration 1

[11:34:44] [INFO]   |-- 📋 Rewrite complete (0 failed) [721.6s]

[11:34:44] [INFO] 🎉 Pipeline complete — 25 records processed, 0 total failures

Out[9]:

	biography	biography_rewritten	utility_score	leakage_mass	weighted_leakage_rate	any_high_leaked	needs_human_review
0	Bobby Watford, a 40‑year‑old Mexican veterinar...	Ethan Kline, a 40‑year‑old Mexican veterinaria...	0.918182	0.0	0.0	False	False
1	Jodi Allison, 36, lives at 204 Bluegrass in Cl...	Claire Harper, in her mid‑30s, lives at 317 Ma...	0.99375	0.0	0.0	False	False
2	James Mills is a 69‑year‑old paramedic who liv...	Victor Harper is a man in his late 60s who wor...	0.966667	0.0	0.0	False	False
3	Nancy Burton is a 21‑year‑old cashier who live...	Emily Hawkins is in her early twenties and wor...	0.98	0.297	0.026757	False	False
4	Cheryl Gray is a 33‑year‑old historian living ...	Denise Bennett is a historian in her late 30s,...	0.933333	0.0	0.0	False	False

In [10]:

result.dataframe[["biography_rewritten", "utility_score", "leakage_mass", "needs_human_review"]].head()

result.dataframe[["biography_rewritten", "utility_score", "leakage_mass", "needs_human_review"]].head()

Out[10]:

	biography_rewritten	utility_score	leakage_mass	needs_human_review
0	Ethan Kline, a 40‑year‑old Mexican veterinaria...	0.918182	0.0	False
1	Claire Harper, in her mid‑30s, lives at 317 Ma...	0.99375	0.0	False
2	Victor Harper is a man in his late 60s who wor...	0.966667	0.0	False
3	Emily Hawkins is in her early twenties and wor...	0.98	0.297	False
4	Denise Bennett is a historian in her late 30s,...	0.933333	0.0	False

In [11]:

result.trace_dataframe.columns.tolist()

result.trace_dataframe.columns.tolist()

Out[11]:

['biography',
 '_anonymizer_record_id',
 '_raw_detected_entities',
 '_seed_entities',
 '_tag_notation',
 '_seed_validation_candidates',
 '_seed_tagged_text',
 '_validated_entities',
 '_seed_entities_json',
 '_initial_tagged_text',
 '_validated_seed_entities',
 '_augmented_entities',
 '_merged_entities',
 '_merged_tagged_text',
 '_validation_candidates',
 '_detected_entities',
 'biography_with_spans',
 '_latent_entities',
 'final_entities',
 '_entities_by_value',
 '_replacement_map',
 '_replacement_map_source',
 '_domain',
 '_domain_supplement',
 '_domain_supplement_privacy',
 '_sensitivity_disposition',
 '_privacy_qa',
 '_rewrite_disposition_block',
 '_sensitivity_disposition_block',
 '_replacement_map_for_prompt',
 '_full_rewrite',
 'biography_rewritten',
 '_meaning_units',
 '_meaning_units_serialized',
 '_quality_qa',
 '_repair_iterations',
 '_privacy_qa_reanswer',
 '_quality_qa_reanswer',
 '_quality_qa_compare',
 'utility_score',
 'leakage_mass',
 'weighted_leakage_rate',
 'any_high_leaked',
 '_needs_repair',
 '_leaked_privacy_items',
 '_rewritten_text__next',
 'needs_human_review']

🚩 Filter by review flag¶

• Records where automated metrics exceed thresholds are flagged for manual review.
• needs_human_review is threshold-based, so a record can have small nonzero leakage without being flagged.
• Use this to prioritize human attention on the records that need it most.
• See Working with flagged records for guidance on diagnosing and resolving flagged records.

In [12]:

df = result.dataframe
flagged = df[df["needs_human_review"] == True]  # noqa: E712
print(f"{len(flagged)} of {len(df)} records flagged for human review")
flagged.head()

df = result.dataframe flagged = df[df["needs_human_review"] == True] # noqa: E712 print(f"{len(flagged)} of {len(df)} records flagged for human review") flagged.head()

0 of 25 records flagged for human review

Out[12]:

	biography	biography_rewritten	utility_score	leakage_mass	weighted_leakage_rate	any_high_leaked	needs_human_review

🔬 Evaluate (optional)¶

Call evaluate() to run LLM-as-judge scoring on the rewrite result — detection validity and three quality rubrics (privacy, quality, style). Evaluation makes additional LLM calls per record. For larger datasets, evaluate a preview first; this tutorial evaluates all 25 rows to demonstrate the complete workflow. This holistic judge is independent of pipeline leakage scoring, so their assessments may differ. See Evaluation for details.

In [13]:

evaluated = anonymizer.evaluate(result)

evaluated = anonymizer.evaluate(result)

In [14]:

evaluated.display_record(0)

evaluated.display_record(0)

Anonymizer Rewrite Preview (record 0)

Original

Bobby| first_name Watford| last_name, a 40| age‑year‑old Mexican| race_ethnicity veterinarian| occupation living in Denver| city, Colorado| state, grew up on the outskirts of the city and developed a love for animals early on. After graduating from Jefferson High| university, he earned his DVM| degree at the University of Colorado Boulder| university, where he also completed a research stint in wildlife health| field_of_study. Fluent in English| language, Bobby| first_name has always described his upbringing as a blend of small‑town curiosity and the vibrant culture of his community, values that continue to shape his compassionate approach to animal care. Since finishing his training, Bobby| first_name has worked at VCA Animal Hospital| organization_name and later at the Colorado Veterinary Clinic| organization_name, where he now leads a busy mixed‑practice team. He identifies as a Christian Democrat| political_view and often volunteers at local shelters, a habit encouraged by his wife, Maya| first_name, and their two teenage children, Aria| first_name and Leo| first_name. Outside the clinic, Bobby| first_name enjoys hiking the Rockies| place_name with his family and mentoring veterinary students from his alma mater.

Rewritten

Ethan Kline, a 40‑year‑old Mexican veterinarian living in Colorado, grew up in a suburban area and developed a love for animals early on. After graduating from a local high school, he earned his DVM at a state university, where he also completed a research stint in wildlife health. Fluent in English, Ethan has always described his upbringing as a blend of small‑town curiosity and the vibrant culture of his community, values that continue to shape his compassionate approach to animal care. Since finishing his training, Ethan has worked at a local veterinary practice and later at a local veterinary practice, where he now leads a busy mixed‑practice team. He identifies as a Christian Democrat and often volunteers at local shelters, a habit encouraged by his wife, Sofia, and their two teenage children, Nina and Jasper. Outside the clinic, Ethan enjoys hiking the Rockies with his family and mentoring veterinary students from his alma mater.

Scores

Utility: 0.92Leakage: 0.00Weighted Leakage Rate: 0.00Rewrite Needs Review: No

Detection Validity: 0.95

Show 1 flagged detection(s)

Value	Label	Reason
Jefferson High	university	wrong_label: Jefferson High is a high school, not a university.

Judge

privacy: mediumquality: highstyle: high

Entity Disposition

Entity	Label	Sensitivity	Protection
Bobby	first_name	high	replace
Watford	last_name	high	replace
Maya	first_name	high	replace
Aria	first_name	high	replace
Leo	first_name	high	replace
40	age	low	leave_as_is
Mexican	race_ethnicity	low	leave_as_is
veterinarian	occupation	low	leave_as_is
Denver	city	high	remove
Colorado	state	low	leave_as_is
Jefferson High	university	high	generalize
University of Colorado Boulder	university	high	generalize
DVM	degree	low	leave_as_is
wildlife health	field_of_study	low	leave_as_is
English	language	low	leave_as_is
VCA Animal Hospital	organization_name	high	generalize
Colorado Veterinary Clinic	organization_name	high	generalize
Christian Democrat	political_view	low	leave_as_is
Rockies	place_name	low	leave_as_is
married	marital_status	low	leave_as_is
2	number_of_children	low	leave_as_is
male	gender	low	leave_as_is
outskirts of Denver	home_location	high	suppress_inference
mixed practice team lead	current_position	low	leave_as_is

⏭️ Next steps¶

• ⚖️ Rewriting Legal Documents -- rewrite legal text with custom entity labels and domain-specific privacy goals.
• 📊 Evaluation -- learn about the detection validity and rewrite quality judges in detail.
• 🎯 Choosing a Replacement Strategy -- compare Redact, Annotate, Hash, and Substitute if you prefer token-level replacement.
• 🔍 Inspecting Detected Entities -- debug what the detection pipeline found before rewriting.