Overview
Kira Systems is a machine learning platform for contract review and due diligence, founded in 2011 and acquired by Litera in 2021. Its AI identifies, extracts, and analyzes provisions and data points from contracts to accelerate M&A due diligence, real estate, finance, and compliance work. The technology is now a core part of Litera's legal AI strategy.
Our take
Kira, now part of Litera, extracts and analyses contract provisions for due diligence and contract review, sold through annual licensing. Contract analysis at volume is a strong application — the task is finding and comparing clauses across hundreds of agreements, which is exactly what machines do better than tired associates at midnight.
Worth it if
Transactional practices and legal departments reviewing contracts at volume during diligence.
Look elsewhere if
Occasional single-contract review, where the setup effort exceeds reading the document.
✨ Features
Clause & Data Extraction
Automatically detects and pulls 1,400+ clauses and data points across 40+ areas.
Quick Study Training
Train custom machine learning models to find firm-specific provisions without coding.
Built-In Smart Fields
Ships with pre-trained models recognizing hundreds of common contract provisions.
Due Diligence Acceleration
Speeds M&A and contract review, cutting document review time by 50%-90%.
⚖️ Pros & Cons
Pros
Cons
💰 Pricing
Kira (Litera)
Custom
- ML contract analysis platform
- 1,400+ built-in smart fields
- Annual licensing, quote-based
Enterprise / Firm
Custom
- Custom model training (Quick Study)
- High-volume due diligence
- Tailored to seats and usage
Categories
❓ Frequently Asked Questions
How accurate is automated clause extraction?
High on standard provisions in conventionally drafted agreements, and lower on unusual drafting, bespoke language and heavily negotiated clauses — which are frequently the ones that matter. Treat extraction as a first pass that surfaces everything for review, not as a filter that decides what a lawyer sees.
Does it need training on our own documents?
Systems of this kind ship with trained provision models and improve on firm-specific language with additional training. That training is a real project requiring lawyer time to label examples. Budget for it, because out-of-the-box performance on your particular drafting conventions is where expectations most often miss.