// Framework Integrations

Drop VEROQ into any AI stack

Copy-paste examples for every major framework. Search news, pull feeds, find similar coverage, explore story clusters, and query structured data — all with confidence scores and bias detection.

Install

$ pip install langchain-veroq

Full Example

from langchain_openai import ChatOpenAI
from langchain.agents import AgentExecutor, create_openai_tools_agent
from langchain_core.prompts import ChatPromptTemplate
from langchain_veroq import (
    PolarisSearchTool,
    PolarisFeedTool,
    PolarisEntityTool,
    PolarisBriefTool,
    PolarisExtractTool,
    PolarisCompareTool,
    PolarisResearchTool,
    PolarisTrendingTool,
    PolarisVerifyTool,
    PolarisRetriever,
)

# 9 pre-built tools — no boilerplate
tools = [
    PolarisSearchTool(api_key="vq_live_xxx"),
    PolarisFeedTool(api_key="vq_live_xxx"),
    PolarisEntityTool(api_key="vq_live_xxx"),
    PolarisBriefTool(api_key="vq_live_xxx"),
    PolarisExtractTool(api_key="vq_live_xxx"),
    PolarisCompareTool(api_key="vq_live_xxx"),
    PolarisResearchTool(api_key="vq_live_xxx"),
    PolarisTrendingTool(api_key="vq_live_xxx"),
    PolarisVerifyTool(api_key="vq_live_xxx"),
]

prompt = ChatPromptTemplate.from_messages([
    ("system", "You are a news analyst with access to VEROQ intelligence feed."),
    ("human", "{input}"),
    ("placeholder", "{agent_scratchpad}"),
])

llm = ChatOpenAI(model="gpt-5.4", temperature=0)
agent = create_openai_tools_agent(llm, tools, prompt)
executor = AgentExecutor(agent=agent, tools=tools, verbose=True)

result = executor.invoke({"input": "What are the latest developments in AI regulation?"})
print(result["output"])

# ── RAG Retriever ──
# Use PolarisRetriever in any LangChain chain or pipeline
retriever = PolarisRetriever(
    api_key="vq_live_xxx",
    category="ai",
    min_confidence=0.7,
    include_sources="reuters.com,ap.com",
    limit=5,
)
docs = retriever.invoke("AI regulation")
for doc in docs:
    print(doc.page_content[:100])
    print(doc.metadata["confidence"], doc.metadata["bias_score"])

Available Endpoints

GET /api/v1/searchFull-text search with confidence, bias, and source domain filters. Speed tiers: depth=fast | standard | deep.

GET /api/v1/agent-feedLLM-optimized feed, filterable by category, tags, and source domains (include_sources / exclude_sources).

POST /api/v1/extractExtract clean article content from 1-5 URLs. Returns structured text, word count, and metadata.

GET /api/v1/similar/:idFind briefs with similar content by semantic similarity

GET /api/v1/clustersStory clusters grouped by overlapping topics

GET /api/v1/dataStructured data points — funding, revenue, metrics, dates

GET /api/v1/streamReal-time SSE stream of published and trending briefs

POST /api/v1/generate/briefOn-demand intelligence brief generation (API key required)

?include_full_text=trueAdd to /feed or /brief/:id to get raw source article content

Full API Reference Watch Live Stream All Integrations

from langchain_openai import ChatOpenAI from langchain.agents import AgentExecutor, create_openai_tools_agent from langchain_core.prompts import ChatPromptTemplate from langchain_veroq import ( PolarisSearchTool, PolarisFeedTool, PolarisEntityTool, PolarisBriefTool, PolarisExtractTool, PolarisCompareTool, PolarisResearchTool, PolarisTrendingTool, PolarisVerifyTool, PolarisRetriever, ) # 9 pre-built tools — no boilerplate tools = [ PolarisSearchTool(api_key="vq_live_xxx"), PolarisFeedTool(api_key="vq_live_xxx"), PolarisEntityTool(api_key="vq_live_xxx"), PolarisBriefTool(api_key="vq_live_xxx"), PolarisExtractTool(api_key="vq_live_xxx"), PolarisCompareTool(api_key="vq_live_xxx"), PolarisResearchTool(api_key="vq_live_xxx"), PolarisTrendingTool(api_key="vq_live_xxx"), PolarisVerifyTool(api_key="vq_live_xxx"), ] prompt = ChatPromptTemplate.from_messages([ ("system", "You are a news analyst with access to VEROQ intelligence feed."), ("human", "{input}"), ("placeholder", "{agent_scratchpad}"), ]) llm = ChatOpenAI(model="gpt-5.4", temperature=0) agent = create_openai_tools_agent(llm, tools, prompt) executor = AgentExecutor(agent=agent, tools=tools, verbose=True) result = executor.invoke({"input": "What are the latest developments in AI regulation?"}) print(result["output"]) # ── RAG Retriever ── # Use PolarisRetriever in any LangChain chain or pipeline retriever = PolarisRetriever( api_key="vq_live_xxx", category="ai", min_confidence=0.7, include_sources="reuters.com,ap.com", limit=5, ) docs = retriever.invoke("AI regulation") for doc in docs: print(doc.page_content[:100]) print(doc.metadata["confidence"], doc.metadata["bias_score"])