POST
/
search
Search through graph data
curl --request POST \
  --url https://api.praxos.ai/search \
  --header 'API-key: <api-key>' \
  --header 'Content-Type: application/json' \
  --data '{
  "query": "<string>",
  "environment_id": "<string>",
  "top_k": 5
}'
{
  "hits": [
    {
      "score": 123,
      "data": {},
      "sentence": "<string>"
    }
  ]
}

​
Search API

The Search API provides multiple search modalities for querying your Praxos environment data, including our advanced Intelligent Search system that uses AI to orchestrate optimal search strategies.

​
Search Modalities

  • "intelligent" - AI-powered query orchestration with automatic strategy selection
  • Features: Natural language understanding, temporal extraction, multi-strategy execution, analytical operations
  • Best for: Complex queries, natural language input, business intelligence queries

​
Traditional Modalities

  • "fast" - Quick Qdrant vector search for simple lookups
  • "node_vec" - Graph-aware Neo4j search with relationship context
  • "vec_edge" - Edge-centric search for relationship queries
  • "type_vec" - Type-aware search with AI classification

​
Example Requests

curl -X POST "https://api.praxos.ai/search" \
  -H "api-key: prx_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "Find highest transaction amounts in November 2023",
    "environment_id": "env-123",
    "search_modality": "intelligent",
    "max_results": 20,
    "enable_multi_strategy": true,
    "include_graph_context": true
  }'

curl -X POST "https://api.praxos.ai/search" \
  -H "api-key: prx_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "customer complaints",
    "environment_id": "env-123", 
    "search_modality": "fast",
    "top_k": 10,
    "node_type": "schema:Review"
  }'

​
Request Parameters

​
Core Parameters

  • query (string, required): Search query text
  • environment_id (string, required): Target environment ID
  • search_modality (string): Search strategy (“intelligent”, “fast”, “node_vec”, “vec_edge”, “type_vec”)

​
Intelligent Search Parameters

  • max_results (integer): Number of results (default: 20)
  • enable_multi_strategy (boolean): Enable multiple strategies (default: true)
  • force_strategy (string): Force specific strategy (optional)
  • anchors (array): Anchor constraints for graph filtering

​
Filtering Parameters

  • source_id (string): Limit search to specific source
  • node_type (string): Filter by node type
  • node_kind (string): Filter by node kind (entity, literal, edge_sentence)
  • temporal_filter (object): Time-based filtering
  • include_graph_context (boolean): Include relationship data

​
Response Structure

​
Intelligent Search Response

{
  "hits": [
    {
      "score": 0.95,
      "sentence": "High-value transaction of $15,750 processed in November 2023",
      "data": {...},
      "type": "FinancialTransaction",
      "graph_context": {...}
    }
  ],
  "intelligent_analysis": {
    "intent": "analytical_query",
    "strategies_used": ["node_vec", "analytical"],
    "temporal_anchors": [
      {"type": "MonthOfYearType", "value": "11"},
      {"type": "YearNumberType", "value": "2023"}
    ],
    "relevant_types": ["FinancialTransaction", "CurrencyAmountType"],
    "execution_time_ms": 245
  },
  "toolset_analysis": [
    {
      "operation": "find_maximum",
      "field": "amount", 
      "max_value": 15750.00,
      "max_item": "Transaction #TXN-2023-11-456",
      "total_items_analyzed": 1247
    }
  ]
}

​
Traditional Search Response

{
  "hits": [
    {
      "score": 0.85,
      "data": {...},
      "sentence": "Generated description text"
    }
  ],
  "total": 150,
  "execution_time_ms": 45
}

​
Analytical Operations

When using intelligent search, the system automatically performs relevant analytical operations:
OperationTrigger KeywordsResponse Field
find_maximum”highest”, “maximum”, “top”, “largest”max_value, max_item
find_minimum”lowest”, “minimum”, “smallest”min_value, min_item
calculate_average”average”, “mean”average_value, total_items
sum_values”total”, “sum”, “combined”sum_value, items_count
count_items”count”, “how many”, “number of”count, groups

​
Error Responses

{
  "error": {
    "code": "INVALID_SEARCH_MODALITY",
    "message": "Unsupported search modality: invalid_mode",
    "details": {
      "supported_modalities": ["intelligent", "fast", "node_vec", "vec_edge", "type_vec"]
    }
  }
}

​
SDK Integration

Using the Python SDK (recommended): 
from praxos_python import SyncClient

client = SyncClient(api_key="prx_your_api_key_here")
environment = client.get_environment(id="env-123")

# Intelligent search
results = environment.intelligent_search(
    query="Find highest transaction amounts in November 2023"
)

# Traditional search  
results = environment.search(
    query="customer complaints",
    search_modality="fast",
    top_k=10
)

Authorizations

​
API-key
string
header
required

API key for authentication

Body

application/json
​
query
string
required

The search query text

​
environment_id
string
required

ID of the environment to search in

​
top_k
integer
default:5

Number of results to return

Required range: 1 <= x <= 100

Response

Successful search results

​
hits
object[]