AI for
SensorIndustrialMedicalFinancialIoTOperationalMachineMarketSensor
data.
Chronicle reads sensor streams and text together. Finetune it on your own data from $5, see proof it beats the base model, and ship it as a private endpoint — all through one API.
Try Chronicle.
Our first foundation model, built on published research — read the paper, run the code, pull the weights.
Preprint2026
Chronicle: A Multimodal Foundation Model for Joint Language and Time Series Understanding
Your data.
Your model.
Your endpoint.
Upload to deployed in minutes. You only ever pay for a model that measurably beats the base — everything else refunds itself.
Send a signal, get an answer.
Embeddings are OpenAI-compatible; forecasting and finetuning are three calls each. Read the docs
python
from openai import OpenAI
client = OpenAI(api_key="iai_...", base_url="https://inertialai.com/api/v1")
def embed(time_series, text=""):
resp = client.embeddings.create(
model="inertial-embed-alpha",
input={"time_series": time_series, "text": text},
)
return resp.data[0].embedding
query_emb = embed(time_series=[0.42, 1.87, 0.95, 2.31, 0.18])
sensor_embs = [
embed(time_series=r["values"], text=r["label"])
for r in sensor_log
]
ranked = sorted(sensor_embs, key=lambda e: cosine_sim(e, query_emb), reverse=True)
results = ranked[:5]Start with $5 of free usage.
Enough to embed, forecast, and train your first model. No subscription, no minimum spend, no sales call.