Skip to main content

​
What is Synkro?

Synkro transforms your policies, handbooks, and documentation into high-quality training datasets for LLMs. Instead of manually crafting training examples, you provide your source documents and Synkro generates diverse, policy-compliant conversations.

​
How It Works

1

Document

Upload your policy (PDF, DOCX, TXT, URL, or raw text)
2

Logic Map

Synkro extracts structured rules with conditions, actions, and dependencies
3

Scenarios

Diverse test cases generated: positive, negative, edge cases
4

Responses

Policy-compliant assistant responses created for each scenario
5

Grade & Refine

Responses evaluated and auto-refined until passing quality checks
6

Dataset

Export training-ready data in your preferred format

​
Key Features

Quality Evaluation

Every response is graded and automatically refined if it fails

Multiple Formats

Conversation, Instruction, Evaluation, and Tool Calling datasets

Coverage Tracking

Track scenario diversity like code coverage, identify gaps

Interactive Editing

Review and edit extracted rules before generation

​
Quick Example

from synkro import create_pipeline
from synkro.examples import EXPENSE_POLICY

pipeline = create_pipeline()
dataset = pipeline.generate(EXPENSE_POLICY, traces=50)
dataset.save("training.jsonl")

​
Next Steps

Quickstart

Generate your first dataset in 5 minutes

API Reference

Explore the full API