Solution · AI & Data
Market Data for AI, Data Warehouses and Alerting Systems
Modern research stacks run on stored time series and live signals. iTick lets you backfill years of K-lines, stream every tick into your time-series database, give large language models real prices instead of stale training data, and alert your team the moment a market moves.
- Scheduled K-line sync plus real-time tick ingestion
- ClickHouse, InfluxDB, TimescaleDB or Parquet
- Live prices and indicators as tools for LLMs and agents
- Breakout, volatility and big-order alerts to chat apps
# WebSocket -> ClickHouse real-time ingestion (batch inserts)
import json, time, websocket, clickhouse_connect # pip install clickhouse-connect
db = clickhouse_connect.get_client(host="localhost")
db.command("""CREATE TABLE IF NOT EXISTS ticks (
sym String, ts DateTime64(3), price Float64, vol Float64
) ENGINE = MergeTree ORDER BY (sym, ts)""")
buf = []
def on_message(ws, msg):
d = json.loads(msg).get("data") or {}
if "ld" in d:
buf.append((d["s"], d.get("t", int(time.time() * 1000)) / 1000, d["ld"], d.get("v", 0)))
if len(buf) >= 500:
db.insert("ticks", buf, column_names=["sym", "ts", "price", "vol"]); buf.clear()
websocket.WebSocketApp("wss://api.itick.org/stock", header={"token": "YOUR_API_TOKEN"},
on_open=lambda ws: ws.send(json.dumps({"ac": "subscribe", "params": "AAPL$US,700$HK", "types": "tick"})),
on_message=on_message).run_forever(ping_interval=20)Three data products
Collection, intelligence and alerting built on one feed.
Market data lake and warehouse
Schedule historical K-line syncs and write WebSocket ticks in real time into ClickHouse, InfluxDB or your lakehouse.
AI research and market LLMs
Provide time-series features and indicators to LLMs for market Q&A, symbol screening and explaining unusual moves.
Price alerts and bots
Breakouts, abnormal volatility and large orders in the book pushed to WeCom, Telegram or DingTalk.
Feature engineering
Returns, volatility, volume profiles and indicators computed from consistent OHLCV and ticks.
Agent-ready
Simple REST calls that fit tool/function calling and MCP-style integrations.
Cross-market signals
Combine US, HK, A-shares, FX and crypto in one model or dashboard.
Data pipeline
From raw feed to insight and action.
- 1
Backfill
REST K-line history on a schedule.
- 2
Stream
WebSocket ticks written in batches to your TSDB.
- 3
Model
Features, indicators and LLM tools read from storage and live quotes.
- 4
Act
Alerts to chat apps, dashboards or trading systems.
Ingest, reason, alert
Three short samples for the three building blocks.
- Get a free API token
- Create your tick table and start ingesting
- Add LLM tools and alert rules on top
# WebSocket -> ClickHouse real-time ingestion (batch inserts)
import json, time, websocket, clickhouse_connect # pip install clickhouse-connect
db = clickhouse_connect.get_client(host="localhost")
db.command("""CREATE TABLE IF NOT EXISTS ticks (
sym String, ts DateTime64(3), price Float64, vol Float64
) ENGINE = MergeTree ORDER BY (sym, ts)""")
buf = []
def on_message(ws, msg):
d = json.loads(msg).get("data") or {}
if "ld" in d:
buf.append((d["s"], d.get("t", int(time.time() * 1000)) / 1000, d["ld"], d.get("v", 0)))
if len(buf) >= 500:
db.insert("ticks", buf, column_names=["sym", "ts", "price", "vol"]); buf.clear()
websocket.WebSocketApp("wss://api.itick.org/stock", header={"token": "YOUR_API_TOKEN"},
on_open=lambda ws: ws.send(json.dumps({"ac": "subscribe", "params": "AAPL$US,700$HK", "types": "tick"})),
on_message=on_message).run_forever(ping_interval=20)AI and data FAQ
Which databases work best for market data?
Columnar and time-series stores such as ClickHouse, InfluxDB or TimescaleDB handle tick volumes well. Batch inserts of a few hundred rows keep ingestion efficient.
Can I use iTick with LLMs and AI agents?
Yes. Expose quote and K-line endpoints as tools so the model can fetch live prices instead of relying on training data. iTick also offers its own AI financial agent.
Which chat apps can receive alerts?
Any app with a webhook or bot API, including Telegram, WeCom (Enterprise WeChat) and DingTalk.
May I store the data?
Storing data for your internal analysis is a common use; check your plan for redistribution terms if stored data is shown to third parties.
Give your models and data stack live markets
Start with a free key and stream your first ticks into storage today.