Match Lookup

Resolve a single company or person to their record in the Landbase graph.

This skill is for single-record lookups only. For matching a full dataset of records in bulk, use workflow match via dataset-pipeline (upload → onboard → match → enrich).

Two subcommands — match company and match person — each with exact and fuzzy modes depending on what information is available.

Quick dispatch

Match company

Exact lookup (preferred when you have a domain or LinkedIn URL)

# By domain — most reliable
landbase-cli match company --website=anthropic.com

# By LinkedIn URL
landbase-cli match company --linkedin-url=https://www.linkedin.com/company/anthropic

# By Landbase company ID
landbase-cli match company --company-id=12345

# Narrow by country when ID is ambiguous
landbase-cli match company --company-id=12345 --hq-country=US

Fuzzy lookup (when you only have a name)

# By name alone
landbase-cli match company --name=Anthropic

# Name + location (improves precision for common names)
landbase-cli match company --name="Landbase" --hq-location-phrase="Los Angeles"

# Name only, skip website resolution
landbase-cli match company --name=Acme --no-name-to-website

When to use --no-name-to-website: when the company name is generic or shared across many entities and you don’t want Landbase to attempt website inference — it prevents false positives at the cost of lower match rate.

Match person

Exact lookup

# By LinkedIn URL — most reliable for people
landbase-cli match person --linkedin-url=https://www.linkedin.com/in/danielsaks

# By email
landbase-cli match person --email=daniel@landbase.com

# By Landbase person ID
landbase-cli match person --person-id=67890

Name-based lookup

# Name + company (recommended — reduces ambiguity)
landbase-cli match person --first-name=Daniel --last-name=Saks --company-website=landbase.com

# Name only
landbase-cli match person --first-name=Daniel --last-name=Saks

# Expand aliases for common nicknames (Bill → William, Bob → Robert, etc.)
landbase-cli match person --first-name=Bill --last-name=Gates --expand-first-name-aliases

When to use --expand-first-name-aliases: when the name provided might be a nickname or shortened form. Without it, “Bill Gates” won’t match a record stored as “William Gates.”

Chaining: company → person

Common pattern — look up the company first to get a reliable domain, then find a person there:

# Step 1: resolve the company
landbase-cli match company --name="Acme Corp" --hq-location-phrase="New York"

# Step 2: use the resolved domain to find a person
landbase-cli match person --first-name=Jane --last-name=Smith --company-website=acme.com

This is more reliable than passing a fuzzy company name to match person directly.

When match fails

If match company or match person returns no result:

1. Try an alternate identifier — if name matching failed, try domain or LinkedIn URL
2. Loosen the query — remove --hq-location-phrase or --hq-country constraints
3. Try --expand-first-name-aliases for person lookups with short first names
4. Fall back to search — landbase-cli search "find [company/person description]" may surface the record via the NL agent

Don’t report a match failure without trying at least one alternate approach.

Red Flags — STOP

• About to loop this command over many records → use dataset-pipeline (workflow match) for bulk matching instead
• Using match company --name for a very generic name (e.g. “Services LLC”) without a location → add --hq-location-phrase first
• Using match person with only a first name → always provide last name at minimum
• Reporting “not found” after one attempt → try an alternate identifier before giving up