Everything you need from
training to deployment
Train in Studio, import from anywhere, deploy instantly, and let anyone query your models.
Studio
Write Python in the browser. Train any model, then export to ML-Dash with one line.
- In-browser notebook with full Python support
- Export sklearn, XGBoost, LightGBM, and CatBoost via SDK
- Export trained models with mldash.export()
- Pre-built templates for common workflows
from sklearn.ensemble import RandomForestClassifier
import mldash
# Train your model
model = RandomForestClassifier(n_estimators=100)
model.fit(X_train, y_train)
# Export to ML-Dash
mldash.export(
model=model,
name="churn-predictor",
target={"churn": "categorical"},
)SDK Export
Train in your own environment. Export your model to ML-Dash with one line of code.
- pip install mldash-sdk to get started
- mldash.export() bundles model + metadata + artifacts
- Auto-detection of task type, framework, features, and class labels
- Works from Colab, Jupyter, VS Code, or any Python environment
import mldash
# Train your model as usual
model.fit(X_train, y_train)
# Export to ML-Dash
mldash.export(
model=model,
name="churn-predictor",
target={"churn": "categorical"},
)
# Your API endpoint is ready instantlyAPI Endpoints
Every trained model gets a REST endpoint. Authenticate, predict, integrate.
- POST /api/v1/predict/{model-slug}/ for single predictions
- API key authentication with X-API-Key header
- Rate limiting by subscription tier (Free / Plus / Pro)
- 48ms average response time
import requests
resp = requests.post(
"https://api.mldash.com/v1/predict/sentiment-analyzer",
headers={"X-API-Key": "mldash_..."},
json={"text": "Great product!"}
)
# {"prediction": "positive", "confidence": 0.94}Chat Interface
Let non-technical users ask questions and get predictions through conversation.
- Describe your data conversationally and the LLM extracts features
- Ask "why?" and get SHAP/LIME explanations narrated by an LLM
- Each model becomes a named agent with its own persona
- Schema-aware validation with graceful error recovery
Based on the churn model, this customer has an 87% probability of churning.
Key factors: high support ticket count (top 15%), moderate tenure.
Explore & Sharing
Publish models to the community. Users discover, try, and bookmark public models.
- Set models to Public with one click
- Browse and search public models on the Explore page
- Try any public model with live predictions
- Bookmark favorite models and datasets
Based on the churn model, this customer has an 87% probability of churning.
Key factors: high support ticket count (top 15%), moderate tenure.
Dataset Management
Upload, preview, override column types, and manage schemas before training.
- CSV and Excel upload (up to 100MB)
- Automatic column type detection with manual overrides
- Preview data with automatic schema detection
- Share datasets publicly on the Explore page
Ready to deploy your first model?
Go from a trained model to a production API in under 5 minutes.