lighthouse-private-markets-sdk 0.1.6

Python SDK for Lighthouse Private Markets API with PostgREST-style querying

Lighthouse API Client (Python)

Lighthouse is an AI-powered platform that automates the entire venture capital deal flow cycle—from sourcing to due diligence. Our platform empowers investors with real-time insights and seamless workflows, helping VCs identify and evaluate high-potential startups more efficiently. Visit us at https://trylighthouse.vc to learn more.

Installation

Using pip:

pip install lighthouse-private-markets-sdk

Quick Start

Importing and Initialization

from lighthouse import Lighthouse

# Initialize the client
lighthouse = Lighthouse('your-api-key')

Fetching Startups

def get_startups():
    try:
        response = lighthouse.from_('startups').select('*').limit(10).execute()
        if response.error:
            raise Exception(response.error)
        print("Startups:", response.data)
    except Exception as err:
        print("Error:", err)

get_startups()

Common Query Examples

Select Single Record

# Get a specific startup
response = lighthouse.from_('startups').select('*').eq('id', 123).single().execute()

Filter Records

# Get all startups valued over $1M
response = lighthouse.from_('startups').select('name, valuation').gte('valuation', 1000000).order('valuation', desc=True).execute()

Select with Relationships

# Get startups with their founders
response = lighthouse.from_('startups').select('name, founders (name, role)').execute()

Pagination

# Get 10 records with offset
response = lighthouse.from_('startups').select('*').range(0, 9).execute()

Full Text Search

# Search startups by name
response = lighthouse.from_('startups').select('*').fts('name', 'tech').execute()

f

Query Building

The Lighthouse client uses a builder pattern for constructing queries. The typical flow is:

1. Start with from_() to specify the table
2. Use select() to specify columns
3. Add filters and modifiers
4. Call execute() to run the query

Core Classes

Lighthouse

The main client class.

• from_(table: str): Start building a query for the specified table

LighthouseSelectQuery

Intermediate query builder after from_() but before select().

• select(*columns: str): Choose specific columns to return

LighthouseFilterQuery

The main query builder with all filter methods.

Available Methods

Basic Filters

• select(*columns: str): Choose specific fields to return
• single(): Request a single record
• maybe_single(): Retrieve at most one row from the result
• eq(column: str, value: Any): Equal to filter
• neq(column: str, value: Any): Not equal to filter
• gt(column: str, value: Any): Greater than filter
• lt(column: str, value: Any): Less than filter
• gte(column: str, value: Any): Greater than or equal to filter
• lte(column: str, value: Any): Less than or equal to filter
• is_(column: str, value: Any): 'IS' filter
• in_(column: str, values: Iterable[Any]): 'IN' filter for array of values

Text Search

• like(column: str, pattern: str): Pattern matching (case sensitive)
• ilike(column: str, pattern: str): Pattern matching (case insensitive)
• like_all_of(column: str, pattern: str): Match all patterns
• like_any_of(column: str, pattern: str): Match any pattern
• ilike_all_of(column: str, pattern: str): Match all patterns (case insensitive)
• ilike_any_of(column: str, pattern: str): Match any pattern (case insensitive)

Full Text Search

• fts(column: str, query: Any): Full text search
• plfts(column: str, query: Any): Plain full text search
• phfts(column: str, query: Any): Phrase full text search
• wfts(column: str, query: Any): Websearch full text search

Array/JSON Operations

• contains(column: str, value: Union[Iterable[Any], str, Dict[Any, Any]]): Check if array/json contains value
• contained_by(column: str, value: Union[Iterable[Any], str, Dict[Any, Any]]): Check if contained by value
• overlaps(column: str, values: Iterable[Any]): Overlaps with values
• cs(column: str, values: Iterable[Any]): Contains filter (shorthand)
• cd(column: str, values: Iterable[Any]): Contained by filter (shorthand)
• ov(column: str, value: Iterable[Any]): Overlaps filter (shorthand)

Range Operations

• sl(column: str, range: Tuple[int, int]): Strictly left of range
• sr(column: str, range: Tuple[int, int]): Strictly right of range
• nxl(column: str, range: Tuple[int, int]): Not extending left of range
• nxr(column: str, range: Tuple[int, int]): Not extending right of range
• adj(column: str, range: Tuple[int, int]): Adjacent to range
• range_gt(column: str, range: Tuple[int, int]): Greater than range
• range_gte(column: str, range: Tuple[int, int]): Greater than or equal to range
• range_lt(column: str, range: Tuple[int, int]): Less than range
• range_lte(column: str, range: Tuple[int, int]): Less than or equal to range
• range_adjacent(column: str, range: Tuple[int, int]): Adjacent to range

Result Modification

• order(column: str, desc: bool = False, nullsfirst: bool = False, foreign_table: Optional[str] = None): Sort results
• limit(size: int, foreign_table: Optional[str] = None): Limit number of rows
• range(start: int, end: int, foreign_table: Optional[str] = None): Get a range of rows
• offset(size: int): Set the starting row index

Logical Operations

• not_(): Negate the next filter
• or_(filters: str, reference_table: Optional[str] = None): OR condition
• match(query: Dict[str, Any]): Match multiple columns
• filter(column: str, operator: str, criteria: str): Apply a custom filter

Output Format

• csv(): Return results as CSV
• explain(analyze: bool = False, verbose: bool = False, settings: bool = False, buffers: bool = False, wal: bool = False, format: Literal["text", "json"] = "text"): Get query explanation

Execution

• execute() -> APIResponse[_ReturnT]: Execute the query and return results

Examples

# Complex query example
response = (lighthouse.from_('startups')
    .select('name, valuation, founders')
    .gte('valuation', 1000000)
    .ilike('name', '%tech%')
    .order('valuation', desc=True)
    .limit(10)
    .execute())

# Full text search
response = (lighthouse.from_('startups')
    .select('*')
    .fts('description', 'artificial intelligence')
    .execute())

# Array operations
response = (lighthouse.from_('startups')
    .select('*')
    .contains('tags', ['AI', 'ML'])
    .execute())

# Range operations
response = (lighthouse.from_('funding_rounds')
    .select('*')
    .range_gte('amount', (1000000, 5000000))
    .execute())

Error Handling

try:
    response = lighthouse.from_('startups').select('*').execute()
except Exception as e:
    print(e)


# Process your data
print(response.data)

Best Practices

1. Always check for errors before using data
2. Use try/except blocks for error handling
3. Limit your selections to required fields only
4. Use pagination for large datasets
5. Cache results when appropriate

Support

• 📧 Email: [e-mail]

License

MIT License - see the LICENSE file for details.