Velari
Velari is an AI deal origination platform for private equity that uses autonomous agents to research companies against a fund's specific investment thesis, uncovering off-market acquisition targets that traditional databases miss. The platform consolidates target identification, validation, and enrichment into a single workflow, with a research pipeline that achieved 94.4% per-criterion accuracy in a 50-company benchmark against frontier LLM baselines.
- AI Agents
- Artificial Intelligence
- Data & Analytics
- Software Only
- Financial Technology
Funding
Founders
Founder details are not available yet.
Product
Problem
Traditional deal sourcing tools rely on structured databases and manual screening, which miss the vast, constantly changing universe of off-market companies that meet highly specific investment criteria. This fragmented approach forces private equity firms and search funds to use multiple disconnected tools, resulting in incomplete target lists and inefficient workflows.
Solution
Velari uses AI agents to solve the core retrieval problem behind exhaustive company search, scanning across massive, unstructured data sources to surface off-market targets that current tools overlook. The platform consolidates several previously separate tools into one system, producing higher-quality outputs while reducing the cost and complexity of building this capability in-house. Velari enriches and validates potential targets and contacts, enabling users to identify, validate, and act on acquisition opportunities with greater speed and precision. The platform's research pipeline reads each company's website, financial filings, ownership records, and local news to establish what a company actually does, not what its industry code claims.
Target Audience
Primary customers are search funds, private equity firms, and M&A brokers who need to identify and validate off-market acquisition targets and relevant business contacts.
Features
- AI-powered agents that continuously search unstructured public data sources to discover off-market companies matching complex thesis criteria
- Consolidated workflow that replaces multiple standalone tools for target identification, validation, and enrichment
- Company and contact enrichment using public registries, corporate websites, and commercial data providers
- Custom scoring rules and tags that let users configure workflows around specific investment theses
- Data processing infrastructure designed for B2B context, with customer data ownership and confidentiality protections
- Multi-stage filtering pipeline that narrows broad universes (e.g., ~151,300 companies to ~110 high-conviction targets) through business model, financial, and ownership analysis
- Benchmark-validated research pipeline achieving 94.4% per-criterion accuracy versus 87.5% (GPT-5.5) and 82.7% (Claude Opus 4.7) on a 21-criterion, 50-company screening benchmark