Expert Witness Directory

Pull expert witness and medico-legal directory records from United States and Australian sources into one deduped list. Each person gets one record, however many directories list them, with specialties mapped to one shared taxonomy and a published email only where a source prints one.

Coverage today: United States and Australia

• The actor returns records for the United States and Australia only. It reads 18 sources: 7 in the US and 11 in Australia, counting ABIME once in the US.
• ABIME lists experts in several countries. The actor reads every ABIME listing and returns only its US and Australian records.
• UK and Canadian records are not offered. A country value other than US or AU stops the run with an error, and no UK or Canadian record is ever written to the dataset.
• The table below lists every US and Australian source with its status. "Implemented" sources parse live. "Planned (stub)" sources are recorded but not read yet. "Not covered" sources sit behind a bot challenge or a robots.txt block, and the actor does not read them.

What it does

• Reads open directory pages, sitemaps, JSON endpoints, and PDFs.
• Filters by specialty, state, country, and source.
• Merges the same person across directories. A doctor on six panels comes back as one record with six sources.
• Maps every source's own specialty tags to one taxonomy, so "Forensic Psychiatry" and "Psychiatrist, general adult" filter together.
• Keeps records between runs in a named key-value store in your account. In feed mode a run returns only the experts that are new, changed, or removed since the last run.
• Optionally checks US physicians and psychologists against the free NPPES monthly file and adds the NPI.

Sources and their terms

Read this table before you use the data. It gives the terms posture of every source as each source published it when the actor was built. Several sources publish terms that restrict automated collection, reproduction, commercial use, or soliciting the people they list. The actor reports the terms; it does not decide for you. Make your own call for your use case, and take legal advice where it matters.

Source	Country	Access	Terms posture	Restricts	robots.txt	Email on page	Profiles	Status
ABIME business directory	US, AU	Open (static-html)	none found; /terms-of-use/ returns 404	none found	User-agent: * Disallow: /wp-admin/	some	988 measured (112 Australia region)	Implemented
LA Superior Court psychiatrist and psychologist panels (3 PDFs)	US	Open (pdf)	none	none found	lacourt.org/robots.txt redirects to lacourt.ca.gov/robots.txt, which returned an empty body (measured 2026-09-30); the blob host has none	yes	164 names and 175 emails measured (criminal panel, updated 2026-08-03)	Implemented
SF Bar Register of Experts	US	Open (static-html)	none found (disclaimer and terms pages checked)	none found	robots.txt returns 200 with an empty body, nothing disallowed (measured 2026-09-30)	yes	76 measured 2026-09-30; "over 600" claimed, stale	Implemented
Contra Costa County Bar experts directory	US	Open (static-html)	none found	none found	Disallow: /temp/ and quicksearch and Cloudflare challenge paths; Crawl-delay: 5 (measured 2026-09-30)	yes	17 measured	Implemented
Illinois State Bar expert directory	US	Open (static-html)	none found	none found	disallows named crawlers (Baiduspider, Yandex, and others) only; the User-agent: * groups are commented out or empty (measured 2026-09-30)	yes	37 category pages measured; 30 to 60 listings estimated	Implemented
Louisiana State Bar expert directory PDF	US	Open (pdf)	none found	none found	robots.txt returns a 404 page, so nothing is disallowed (measured 2026-09-30)	yes	about 27 advertisers, 20 emails measured	Implemented
AAIMCO member directory (insurance consultants)	US	Open (static-html)	not located	none found	Yoast default: User-agent: * with an empty Disallow (measured 2026-09-30)	yes	70 measured	Implemented
LACBA expert4law	US	Open (static-html)	LACBA disclaimer and proprietary notice bars reproduction	reproduction	none seen	yes	"more than 1,000" claimed	Planned (stub)
Lexvisio	US	Open (static-html)	none explicit	none found	not restrictive	no	2,153 measured	Planned (stub)
Law.com Experts (ALM)	US	Open (static-html)	none explicit	none found	recorded by the recon, not restrictive for expert pages	no	"15,000+" claimed	Planned (stub)
ExpertPages	US	Disallowed by robots.txt (static-html)	bars republishing	reproduction, automation	User-agent: * Disallow: / (only named search engines allowed)	no	460 measured	Not covered: robots.txt disallows all non-search crawlers and no profile shows an email
JurisPro	US	Open (static-html)	Terms ban unauthorized scraping and separately ban soliciting listed experts by email, phone, text, or fax for other directories, referral services, or similar marketing	automation, solicitation	not restrictive	no	1,962 measured	Planned (stub)
Experts.com	US	Open (static-html)	Terms bar use by scripts, machines, or automated services, and bar using information from the site for uninvited solicitations	automation, solicitation	recorded by the recon	yes	748 measured	Planned (stub)
TrialSmith expert index (names only)	US	Open (static-html)	Terms and Conditions bar reproduction	reproduction	recorded by the recon	no	233,953 name pages measured	Not covered: name and deposition count only, many common-name collisions; kept as a possible verification join, not a record source
SEAK expert witness directory	US	Blocked by bot defense (cloudflare)	no terms page located	none found	allows /members/	unknown	"over 2,000" claimed	Not covered: Cloudflare challenge; no bypass without a separate decision (work order decision 4); Aaron declined a bypass on 2026-09-30
Expert Institute	US	Blocked by bot defense (cloudflare)	not reached	none found	not reached	unknown	"1+ million" claimed	Not covered: Cloudflare challenge; no bypass without a separate decision (work order decision 4); profiles are auto-generated from public sources
American Academy of Psychiatry and the Law	US	Blocked by bot defense (cloudflare)	not reached	none found	not reached (Cloudflare 403 on robots.txt)	unknown	not measured	Not covered: Cloudflare challenge; no bypass without a separate decision (work order decision 4)
Medilaw	AU	Open (static-html)	none found in the sitemap	none found	robots.txt returns HTTP 200 with an empty body (measured 2026-09-30)	panel-only	214 measured 2026-09-30	Implemented
Lex Medicus	AU	Open (static-html)	not located	none found	Disallow: /wp-admin/ and /?html2pdf=* (measured 2026-09-30)	panel-only	184 measured 2026-09-30	Implemented
Australian Specialist Hub	AU	Open (static-html)	not located	none found	User-agent: * Disallow: (everything allowed) (measured 2026-09-30)	panel-only	294 measured 2026-09-30	Implemented
Medico Legal Specialists	AU	Open (static-html)	not located	none found	User-agent: * Disallow: (everything allowed) (measured 2026-09-30)	panel-only	143 measured 2026-09-30	Implemented
Index Medicolegal	AU	Open (static-html)	not located	none found	User-agent: * Disallow: (allow all) (measured 2026-09-30)	panel-only	110 measured 2026-09-30	Implemented
Themis Medico-Legal	AU	Open (static-html)	Terms s.4.3: "You must not copy, reproduce, modify, distribute, display, or exploit any content without our prior written consent."	reproduction, commercial-use	Disallow: /wp-admin/ (measured 2026-09-30)	panel-only	82 measured	Implemented
IntegrityML	AU	Open (json-api)	Disclaimer: no part "may be reproduced without the specific written permission ... except with appropriate attribution"	reproduction	User-agent: * Disallow: (everything allowed; measured 2026-09-30)	panel-only	57 measured	Implemented
VERIFY Medico-Legal Solutions	AU	Open (static-html)	no scraping, automation, or reproduction clause found	none found	Disallow: /wp-admin/ (measured 2026-09-30)	panel-only	25 measured	Implemented
LIME Medicolegal	AU	Open (static-html)	no scraping or reproduction clause found	none found	sitemap line only (measured 2026-09-30)	panel-only	25 measured	Implemented
Medicolegal Assessments Group directory PDF	AU	Open (pdf)	none found	none found	blocks named downloaders (HTTrack, EmailCollector, and others) with Disallow: /; no User-agent: * group (measured 2026-09-30)	no	about 486 measured by the recon	Implemented
WorkCover WA approved medical specialists	AU	Open (json-api)	none found	none found	recorded by the recon, not restrictive	no	310 measured 2026-09-30 (88 psychiatrists)	Implemented
Specialists Medicolegal Australia	AU	Open (static-html)	not located	none found	Disallow: (allow all)	panel-only	28 measured	Planned (stub)
Medibytes Legal	AU	Open (static-html)	Website terms PDF: "You must not conduct any systematic or automated data collection activities (including without limitation scraping, data mining, data extraction and data harvesting)"; also bars commercial exploitation	automation, commercial-use	Disallow: /wp-admin/	panel-only	22 measured	Planned (stub)
NSW Personal Injury Commission medical assessors	AU	Open (static-html)	none found	none found	recorded by the recon	no	about 180 measured	Planned (stub)
RANZCP Find a Psychiatrist (yourhealthinmind.org)	AU	Open (js)	not checked by the recon	none found	User-agent: * with sitemap only (no disallow)	some	not measured	Planned (stub)
SIRA NSW permanent impairment assessors	AU	Blocked by bot defense (cloudflare)	public government data	none found	not reached	unknown	not measured	Not covered: Cloudflare challenge; no bypass without a separate decision (work order decision 4); manual route for Proofed's own list
WorkSafe Queensland register of permanent impairment trained assessors	AU	Blocked by bot defense (cloudflare)	public government data	none found	not reached	unknown	not measured	Not covered: Cloudflare challenge; no bypass without a separate decision (work order decision 4); manual route for Proofed's own list
WorkSafe Tasmania accredited medical practitioners	AU	Blocked by bot defense (cloudflare)	site requires accepting terms before the list is shown	none found	not reached	unknown	not measured	Not covered: Cloudflare challenge; no bypass without a separate decision (work order decision 4); manual route for Proofed's own list

Sources whose terms restrict commercial use, solicitation, automation, or reproduction

• Themis Medico-Legal: reproduction, commercial-use. Terms s.4.3: "You must not copy, reproduce, modify, distribute, display, or exploit any content without our prior written consent."
• IntegrityML: reproduction. Disclaimer: no part "may be reproduced without the specific written permission ... except with appropriate attribution"
• LACBA expert4law: reproduction. LACBA disclaimer and proprietary notice bars reproduction
• ExpertPages: reproduction, automation. bars republishing
• JurisPro: automation, solicitation. Terms ban unauthorized scraping and separately ban soliciting listed experts by email, phone, text, or fax for other directories, referral services, or similar marketing
• Experts.com: automation, solicitation. Terms bar use by scripts, machines, or automated services, and bar using information from the site for uninvited solicitations
• TrialSmith expert index (names only): reproduction. Terms and Conditions bar reproduction
• Medibytes Legal: automation, commercial-use. Website terms PDF: "You must not conduct any systematic or automated data collection activities (including without limitation scraping, data mining, data extraction and data harvesting)"; also bars commercial exploitation

Sources marked "Blocked by bot defense" sit behind a Cloudflare challenge. This actor does not attempt to bypass bot challenges.

Input

Field	What it does
specialty	Taxonomy codes. A parent code matches its children.
region	State or region as the source prints it.
country	US, AU, or both. Empty returns both.
sources	Directories to read. Empty reads every implemented US and Australian source.
mode	list returns every matching expert. feed returns only new, changed, returned, and removed experts since the last run on the same persistent store.
verifyNppes	Checks US records against the NPPES monthly file.
maxItems	Caps output rows. 0 means no cap.
persistStoreName	Named key-value store that keeps records between runs. Reuse it for feed mode.

Output

One row per deduplicated person per run. A person on three directories is one row with three entries in sources. Main fields: name, post_nominals, specialty_raw, specialty_mapped, jurisdiction, country, firm, practice_site, published_email, email_source, email_type, phone, listing_url, sources, consent_basis, consent_status, npi, nppes_status, change_type.

The actor never guesses an email. published_email is empty unless a source page printed an address.

Change feed

• A record is changed only when its content changes: name, post-nominals, specialty, jurisdiction, country, firm, practice site, email, or phone. A new sitemap date or a new directory that adds nothing is not a change.
• A listing is removed after 2 complete, unfiltered runs do not find it. So removals can first appear on the third run for the same store. A run capped by maxItems, narrowed by a filter, or hit by a fetch error never removes anything.
• A removed row is emitted only when the person has no listing left on any source.

NPPES check

With verifyNppes on, each US record is matched to the NPPES file on last name, first name, and state, and confirmed by specialty. Only a single match counts as verified. More than one is ambiguous. NPPES carries no email, so the check never adds one. The extract covers physicians and psychologists from the newest monthly file (September 2026 when this was written).

Pricing

Pay per event. You pay only for rows written to the dataset.

Event	Price	Charged when
expert-record (base record)	4.00 USD per 1,000	One deduplicated expert row in list mode.
expert-record (change feed row)	4.00 USD per 1,000, the base record rate	One new, changed, returned, or removed row in feed mode. Unchanged experts are not emitted in feed mode and cost nothing.
email-found	10.00 USD per 1,000	The row carries a published email. Removed rows never charge it.
npi-verified	2.00 USD per 1,000	A US row checked against NPPES, with verifyNppes on.

The Actor start event is 0.00005 USD, charged by Apify once per run for each GB of memory.

Consent columns

Two columns are set when a record is first stored and are never recomputed at send time.

• consent_basis comes from the country: US-CAN-SPAM for a US record and AU-SPAM-ACT-inferred for an Australian record.
• consent_status is unreviewed, approved, or blocked, and starts as unreviewed. The actor never approves anything. You review segments and flip them yourself.

These columns are a starting label for your own review, not legal advice.

How duplicates merge

1. The same personal email merges two listings, when the last names agree. A shared inbox such as reception@ never merges people.
2. The same name plus the same phone merges.
3. The same name plus the same practice website merges.
4. The same name plus the same country plus the same specialty family merges, for panels that print only a name, a specialty, and a state. It never merges when the two carry different personal emails or websites, never on an initial-only first name, and once two different people are known to share it, it merges nobody.

Names fold common nicknames first, so Bob Smith and Robert Smith compare as one name. A name alone never merges, because common names belong to different people.

Limitations

• ABIME prints no country or region in its sitemap, so the actor reads every listing page and keeps the US and Australian records after. Expect about 988 page reads for a full ABIME pass at one request per 2 seconds per host.
• Australian panel sources print the panel's contact details, not the doctor's, so those records carry a name, specialty, and state but no email.
• Nickname folding covers common English short forms only. A name printed with a different spelling on two sources can still come back as two people.

Built by Mamba Labs.