MonceApp v0.1.0

Models Dashboard Paper Architecture Economics CSV CNF Availability Developers API Docs

/csv — Train a SAT Classifier from CSV

Send inline CSV data, get a trained Snake model back. Powered by snakebatch.aws.monce.ai — distributed Lambda, handles 150K+ rows.

Via Chat (model calls it for you)

# Just tell the model what you want
curl -sX POST https://monceapp.aws.monce.ai/v1/chat \
  -F "model_id=eu.anthropic.claude-haiku-4-5-20251001-v1:0" \
  -F "message=Train on this: name,size,label
alice,small,good
bob,large,bad
charlie,small,good" | jq .reply

# Model calls snake_csv tool internally, returns natural answer

Via API (direct)

# POST to snakebatch directly
curl -sX POST https://snakebatch.aws.monce.ai/v6/train \
  -H "Content-Type: application/json" \
  -d '{
    "data": [
      {"name":"alice","size":"small","label":"good"},
      {"name":"bob","size":"large","label":"bad"}
    ],
    "config": {"target_index":"label","n_layers":25,"bucket":16,"noise":0.25}
  }'

# → {"model_id": "snake-abc-123", "wall_clock_ms": 1200}

Via Python SDK

from monceai import Snake

model = Snake([
    {"name":"alice","size":"small","label":"good"},
    {"name":"bob","size":"large","label":"bad"},
    {"name":"charlie","size":"small","label":"good"},
], target_index="label", mode="fast")

model.get_prediction({"name":"dave","size":"small"})  # → "good"
model.get_audit({"name":"dave","size":"small"})        # → SAT reasoning trace

Constraints