Custom data connectors, desktop app for all seats, SAML/SCIM SSO, private VPC deployment, pooled credits, invoice billing, dedicated support, and SLA for teams deploying ML at scale.
Install OctOpus on your machine and train models locally — your data never leaves your device. Available on Team and Enterprise plans.
OctOpusCustom API
Programmatic access
Custom API
Integrate OctOpus into your stack. Train models, trigger predictions, and fetch results from your own app. Available on Pro, Pro+, Team, and Enterprise.
OctOpusCustom Connectors
Enterprise data sources
Custom Connectors
Plug OctOpus directly into your warehouse, data lake, or internal system. Secure, private, and tailored to your infrastructure. Enterprise only.
OctOpus
Automated data science for enterprise teams
or continue with email
Use email & password
API Keys
Connect your own API keys. Your keys are stored securely and never shared. They're used to power your ML experiments.
Kimi / MoonshotNot set
OpenAI (GPT-4)Not set
Anthropic (Claude)Not set
Google (Gemini)Not set
Amazon Bedrock (Claude)Not set
✓ Saved
Keys are sent over HTTPS and stored encrypted on our servers. You can update or remove them at any time.
Ephemeral storage detected — runs, models & dashboards will be LOST on every container restart.
Mount /data
+ set STORAGE_DIR=/data.
Warehouse · database · or drop a file (CSV · Excel · JSON · ZIP · PDF)
Drop test.csv to predict
model ready · drop test set here
BI · Featured
Recommended
YC companies — real alumni
5,866 startups · industries, team size, batches, status
Geo · WMO
Live API
WMO world weather — 500 cities
Lat/lon, forecasts & climate normals · world map ready
Healthcare & life sciences
Healthcare
Hospital readmissions
4,200 encounters · 30-day risk
Research
Clinical trial outcomes
2,400 subjects · 3 arms, 12 sites
Biology
Gene expression — tumor subtype
600 samples × 40 genes
Environment & hydrology
Hydrology
River discharge — daily
3 years · rain, snowmelt, dam release
Hydrology
Water quality stations
2,200 samples · good / moderate / poor
Ecology
Species richness survey
1,300 plots · habitat & climate
Agriculture
Crop yields
2,600 field-seasons · 5 crops
Business & industry
Retail
Retail orders
8,000 order lines · demand
Finance
Loan defaults
5,000 loans · risk scoring
Energy
Grid load — hourly
4,368 hours · temp-driven
Manufacturing
Machine failures
5,000 machines · sensor data
Education
Student exam scores
2,000 students
Audio & language AI
Audio · Real speech
Spoken digits — real recordings
60 real voice clips · zero…nine · ASR-ready
Audio · TTS corpus
Wolof speech — real TTS corpus
Real recordings + transcripts · 2 native speakers · CC BY 4.0
Audio
Machine sounds — audio ML
80 recordings · alarm / hum / bell / static
ASR · NLP
Support call transcripts
900 calls · intent from the text
LLM · Real data
Dolly instruction pairs — real
1,200 human-written · fine-tune ready
LLM
LLM fine-tune — support assistant
600 instruction → answer pairs
Classics
Climatology
WMO monthly normals
3,792 city-months · seasonality
Classification
Titanic
891 rows
Multiclass
Iris
150 rows
Multiclass
Palmer penguins
344 rows
Classification
Telco churn
7,043 rows
Regression
House prices
1,460 rows
Regression
Diamonds
53,940 rows
Regression
Auto MPG
398 rows
Time-series
Airline passengers
144 rows
Connect a warehousePro+ & Enterprise
More import options
Sources
Status--
Experiments0
Best--
Run--
Research Plan
Home
Demo
Data
Chat
Plan
Exp
Dashboard
Model
Projects
Advanced
Docs
Descartes ↗
Advanced tools
Power tools, diagnostics, and outputs. Open one when you need depth.
Idle
Connect your data to get started.
Import a dataset or load a sample, then ask a business question in plain language.
Prefer it hands-on?Watch OctOpus run a real training session, live in this workspace.
Step-by-step walkthroughThe full journey — data in, model out — as a guided replay you can step through.
Trusted by
train.py
workspace›train.pyEditable — Ctrl/⌘ S save · Ctrl/⌘ ⏎ run
OUTPUT
SavedLn 1, Col 1Spaces: 4UTF-8Python
◆ Research plan (program.md)◆ Your planLIVEEditable by you — save the plan, then run from this workspace whenever you want.
Research plan
Plan before running
Approve the plan in chat and I’ll write program.md before the first experiment. After that, you can edit any line here, save it, and run again from the same workspace.
Edit the plan freely. Save it to update program.md, or use
Save + run to continue from the edited workspace immediately.
During a live run, saved edits apply to the next experiment.
Training runstarting…
Profile
Plan
Train
Validate
Finalize
Getting ready…
▼Research Plan
Load a dataset and ask a business question. I'll profile it, plan an analysis, run it, and explain what to do.
Working…View live
Listening — speak naturallyTap to end
Connecting…
Tools
Integrations
Bring OctOpus into your editor, notebook and stack.
Available
VS Code
Editor
Profile CSVs, launch runs and stream logs without leaving your editor.
Available
Cursor
Editor
The OctOpus extension and MCP server, right inside Cursor.
Available
Claude
Assistant · MCP
Add the OctOpus MCP server to Claude Desktop or Claude Code.
Available
Jupyter
Notebook
Pull trained models and predictions straight into a notebook.
Available
Python SDK & API
Developer
Automate runs and scoring from your own code.
Available
Dataiku
Data platform
Native DSS plugin — Train and Score recipes in your Flow, SDK bundled, no PyPI needed. Or paste the one-cell recipe.
Soon
GitHub
Source control
Trigger runs on push and post results back on pull requests.
Soon
Slack
Alerts
Run, champion and deploy notifications in your channels.
