Developers - Built for developers. Designed for scale.

Catalyzed Data

Query across your private datasets and public marketplace data with standard SQL. Vector search and joins in a single query.

500+ connectors Automatic indexing Petabyte scale Billions of vectors

import requests

API_URL = "YOUR_API_URL"
API_KEY = "YOUR_API_KEY" # Find patents similar to a concept and join with claims
response = requests.post(
f"{API_URL}/queries",
    headers={"Authorization": f"Bearer {API_KEY}",
"Content-Type": "application/json" },
    json={"tables": {"patents": "PATENTS_TABLE_ID",
"claims": "CLAIMS_TABLE_ID" },
"sql": """
            SELECT p.title, p.assignee, p._distance, c.claim_text
            FROM knn_search('patents', 'abstract_embedding',
text_to_embedding('machine learning image classification'), 5) p
            JOIN claims c ON p.patent_id = c.patent_id
            WHERE c.claim_type = 'independent'
            ORDER BY p._distance
        """ }
)

result = response.json()
# {
#   "rows": [
# {"title": "Deep Learning System for Radiological Image Analysis", "_distance": 0.127, ...},
# {"title": "Automated Diagnostic Imaging Pipeline", "_distance": 0.183, ...},
# {"title": "Neural Network for CT Scan Interpretation", "_distance": 0.241, ...}
#   ],
#   "rowCount": 3
# }

const API_URL = "YOUR_API_URL";
const API_KEY = "YOUR_API_KEY";
 const response = await fetch(
`${API_URL}/queries`,
{
method: "POST",
    headers: {"Authorization": `Bearer ${API_KEY}`,
"Content-Type": "application/json" },
    body: JSON.stringify({tables: {
patents: "PATENTS_TABLE_ID",
        claims: "CLAIMS_TABLE_ID" },
sql: `
        SELECT p.title, p.assignee, p._distance, c.claim_text
        FROM knn_search('patents', 'abstract_embedding',
text_to_embedding('machine learning image classification'), 5) p
        JOIN claims c ON p.patent_id = c.patent_id
        WHERE c.claim_type = 'independent'
        ORDER BY p._distance
      ` })
}
);
const result = await response.json();
// {
//   "rows": [
// {"title": "Deep Learning System for Radiological Image Analysis", "_distance": 0.127, ...},
// {"title": "Automated Diagnostic Imaging Pipeline", "_distance": 0.183, ...},
// {"title": "Neural Network for CT Scan Interpretation", "_distance": 0.241, ...}
//   ],
//   "rowCount": 3
// }

# Find patents similar to a concept and join with claims
curl -X POST "YOUR_API_URL/queries" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d @- << 'EOF' {"tables": {"patents": "PATENTS_TABLE_ID", "claims": "CLAIMS_TABLE_ID"},
"sql": "SELECT p.title, p.assignee, p._distance, c.claim_text FROM knn_search('patents', 'abstract_embedding', text_to_embedding('machine learning image classification'), 5) p JOIN claims c ON p.patent_id = c.patent_id WHERE c.claim_type = 'independent' ORDER BY p._distance"
}
EOF
# Response: # {
#   "rows": [
# {"title": "Deep Learning System for Radiological Image Analysis", "_distance": 0.127, ...},
# {"title": "Automated Diagnostic Imaging Pipeline", "_distance": 0.183, ...},
# {"title": "Neural Network for CT Scan Interpretation", "_distance": 0.241, ...}
#   ],
#   "rowCount": 3
# }

Catalyzed Orchestration

Define workflows once, run at scale. Trigger pipelines via API and stream real-time progress events.

API/SDK-defined pipelines Real-time streaming LLM integration Human-in-the-loop

import requests
import json

API_URL = "YOUR_API_URL"
API_KEY = "YOUR_API_KEY"
PIPELINE_ID = "YOUR_PIPELINE_ID" # Trigger pipeline with streaming enabled
response = requests.post(
f"{API_URL}/pipelines/{PIPELINE_ID}/trigger",
    headers={"Authorization": f"Bearer {API_KEY}",
"Content-Type": "application/json",
"Accept": "text/event-stream" },
    json={"input": {"files": {"documents": ["file_abc123", "file_def456"]},
"dataInputs": {"extractionPrompt": "Extract key contract terms"}}},
    stream=True
)
# Stream execution events for line in response.iter_lines():
if line:
        event = json.loads(line.decode().removeprefix("data: "))
        print(f"{event['type']}: {event.get('message', event.get('output', ''))}")
# Events: # {"type": "started", "executionId": "exec_abc123"} # {"type": "progress", "pct": 25, "message": "Processing document 1 of 2..."} # {"type": "progress", "pct": 75, "message": "Extracting terms..."} # {"type": "completed", "output": {"extractedTerms": [...], "confidence": 0.94}}

const API_URL = "YOUR_API_URL";
const API_KEY = "YOUR_API_KEY";
const PIPELINE_ID = "YOUR_PIPELINE_ID";
 const response = await fetch(
`${API_URL}/pipelines/${PIPELINE_ID}/trigger`,
{
method: "POST",
    headers: {"Authorization": `Bearer ${API_KEY}`,
"Content-Type": "application/json",
"Accept": "text/event-stream" },
    body: JSON.stringify({input: {
files: { documents: ["file_abc123", "file_def456"] },
        dataInputs: { extractionPrompt: "Extract key contract terms" }}})
}
);
 const reader = response.body.getReader();
const decoder = new TextDecoder();
while (true) {const { done, value } = await reader.read();
if (done) break;
const event = JSON.parse(decoder.decode(value).replace("data: ", ""));
  console.log(`${event.type}: ${event.message || JSON.stringify(event.output)}`);
} // {"type": "started", "executionId": "exec_abc123"} // {"type": "progress", "pct": 25, "message": "Processing document 1 of 2..."} // {"type": "completed", "output": {"extractedTerms": [...], "confidence": 0.94}}

# Trigger pipeline with streaming enabled
curl -N -X POST "YOUR_API_URL/pipelines/YOUR_PIPELINE_ID/trigger" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Accept: text/event-stream" \
  -d '{"input": {
"files": {"documents": ["file_abc123", "file_def456"]},
      "dataInputs": {"extractionPrompt": "Extract key contract terms"}}}'
# Streamed events: # data: {"type": "started", "executionId": "exec_abc123"} # data: {"type": "progress", "pct": 25, "message": "Processing document 1 of 2..."} # data: {"type": "progress", "pct": 75, "message": "Extracting terms..."} # data: {"type": "completed", "output": {"extractedTerms": [...], "confidence": 0.94}}

Catalyzed Control

Capture expert feedback, measure quality with evaluations, and continuously improve pipelines with AI-generated synthesis.

Feedback capture Evaluation Active learning Continuous improvement

import requests

API_URL = "https://api.catalyzed.ai"
API_KEY = "YOUR_API_KEY"
PIPELINE_ID = "YOUR_PIPELINE_ID" # 1. Capture expert feedback on a pipeline execution
requests.post(
f"{API_URL}/signals",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={"teamId": "team_abc123",
"data": {"type": "comment", "text": "Missing key financial metrics"},
"executionIds": ["exec_xyz789"]
}
)
# 2. Run evaluation against ground truth examples
eval_response = requests.post(
f"{API_URL}/pipelines/{PIPELINE_ID}/evaluate",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={"exampleSetId": "examples_abc",
"evaluatorType": "llm_judge" }
)
evaluation = eval_response.json()
print(f"Evaluation started: {evaluation['evaluationId']}")
# 3. Generate improvements from accumulated feedback
synth = requests.post(
f"{API_URL}/pipelines/{PIPELINE_ID}/synthesize",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={"handlerType": "apg_v1"}
).json()
# 4. Apply proposed changes if synth["status"] == "generated":
    requests.post(f"{API_URL}/synthesis-runs/{synth['synthesisRunId']}/apply",
                  headers={"Authorization": f"Bearer {API_KEY}"})

const API_URL = "https://api.catalyzed.ai";
const API_KEY = "YOUR_API_KEY";
const PIPELINE_ID = "YOUR_PIPELINE_ID";
 await fetch(`${API_URL}/signals`, {
method: "POST",
  headers: { Authorization: `Bearer ${API_KEY}`, "Content-Type": "application/json" },
  body: JSON.stringify({
teamId: "team_abc123",
    data: { type: "comment", text: "Missing key financial metrics" },
    executionIds: ["exec_xyz789"]
})
});
 const evalRes = await fetch(`${API_URL}/pipelines/${PIPELINE_ID}/evaluate`, {
method: "POST",
  headers: { Authorization: `Bearer ${API_KEY}`, "Content-Type": "application/json" },
  body: JSON.stringify({ exampleSetId: "examples_abc", evaluatorType: "llm_judge" })
});
const evaluation = await evalRes.json();
console.log(`Score: ${evaluation.aggregateScore}`);
 const synthRes = await fetch(`${API_URL}/pipelines/${PIPELINE_ID}/synthesize`, {
method: "POST",
  headers: { Authorization: `Bearer ${API_KEY}`, "Content-Type": "application/json" },
  body: JSON.stringify({ handlerType: "apg_v1" })
});
const synthesis = await synthRes.json();
 if (synthesis.status === "generated") {await fetch(`${API_URL}/synthesis-runs/${synthesis.synthesisRunId}/apply`, {
method: "POST",
    headers: { Authorization: `Bearer ${API_KEY}` }});
}

# 1. Capture expert feedback on a pipeline execution
curl -X POST "https://api.catalyzed.ai/signals" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
"teamId": "team_abc123",
    "data": {"type": "comment", "text": "Missing key financial metrics"},
    "executionIds": ["exec_xyz789"]
}'
# 2. Run evaluation against ground truth examples
curl -X POST "https://api.catalyzed.ai/pipelines/YOUR_PIPELINE_ID/evaluate" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"exampleSetId": "examples_abc", "evaluatorType": "llm_judge"}'
# 3. Generate improvements from accumulated feedback
curl -X POST "https://api.catalyzed.ai/pipelines/YOUR_PIPELINE_ID/synthesize" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"handlerType": "apg_v1"}'
# 4. Apply proposed changes (after reviewing)
curl -X POST "https://api.catalyzed.ai/synthesis-runs/SYNTHESIS_RUN_ID/apply" \
  -H "Authorization: Bearer YOUR_API_KEY"

Beta: These APIs are available now. View documentation →

Signals | Evaluations | Synthesis | Full Guide

Built for developers. Designed for scale.

Platform Features

Catalyzed Data

Catalyzed Orchestration

Catalyzed Control

Architecture

Open formats

No vendor lock-in

Your cloud or ours

Security & Compliance

SOC 2 Aligned

HIPAA Ready

GDPR Compliant

Data Residency

Ready to build?