Pecan AI
Pecan is a predictive AI agent that connects to a cloud data warehouse, auto-generates SQL training sets, engineers features, and builds and benchmarks machine-learning models from plain-language business questions. For property operators and REIT analysts this supports tenant churn prediction, rent and demand forecasting, and tenant lifetime-value scoring without an in-house data science team, with predictions pushed back to CRMs and BI dashboards.
Best for: Property and REIT analytics teams that want to build tenant churn, rent or demand forecasting, and lifetime-value models without data scientists.
Last reviewed 6/2/26
- Pricing
- Starter $760/month (2 monthly prediction batches, 500M rows) and Team $1,400/month (10 batches, 2B rows), both billed annually; Business tier custom (5B rows, enterprise deployment). Extra prediction batches $50 each. No setup fees.
- Integrations
- Snowflake, Google BigQuery, Databricks, Amazon Redshift, ClickHouse, MySQL, Microsoft SQL Server, PostgreSQL, Oracle, IBM Db2, Salesforce, HubSpot, Zoho CRM, Amazon S3, Google Cloud Storage
- Property types
- Multifamily, Commercial, Residential
- Geography served
- Global
How Pecan AI uses AI[1][4]
Pecan is a predictive AI agent that connects to a cloud data warehouse, auto-generates SQL training sets, engineers features, and builds and benchmarks machine-learning models from plain-language business questions. For property operators and REIT analysts this supports tenant churn prediction, rent and demand forecasting, and tenant lifetime-value scoring without an in-house data science team, with predictions pushed back to CRMs and BI dashboards.
- • Conversational predictive AI agent turns a plain-language question into a built and validated ML model, handling data prep, feature engineering, and benchmarking automatically.
- • Connects to Snowflake, BigQuery, Redshift, Databricks, and major databases and CRMs, then delivers scheduled predictions back to a database, warehouse, or CRM.
- • Real-estate applications include tenant churn, rent and demand forecasting, and lifetime-value modeling for portfolio teams.
- • Vendor cites aggregate outcomes such as a 12 percent average reduction in churn, a 15 percent improvement in marketing ROAS, and 90 percent of predictions delivered without data science support.
AI type: Predictive machine learning (AutoML) with a conversational LLM agent for model building
API: yes MCP: no
Recognition[2]
Key numbers[2][6]
- • Raised $66M Series C led by Insight Partners (2022), with over $100M raised in 12 months
- • tripled annual recurring revenue year over year
- • vendor cites a 12 percent average reduction in churn, 15 percent improvement in marketing ROAS, and 25 percent reduction in inventory costs
- • Johnson & Johnson reported ROI within weeks on supply-chain forecasting
- • ISO 27001 certified and SOC 2 Type II audited.
Credibility[5][6][2]
- Founders
- Co-founded in 2018 by CEO Zohar Bronfman (two PhDs, background in computational psychology and data science) and Noam Brezis (PhD in computational neuroscience, background in software and data consulting). CTO and SVP R&D Tomer Meron brings 20-plus years at companies including eBay and Google.
- Customers
- Used across fintech, insurance, retail, consumer packaged goods, mobile apps, and consumer services. Named customers include Johnson & Johnson, Kenvue, Little Spoon, Redis, Wild Alaskan, Nanit, Whistle Express Car Wash, and Clearwave Fiber.
- Investors
- Raised a $66M Series C in 2022 led by Insight Partners, with GV (Google Ventures), S-Capital, GGV Capital, Dell Technologies Capital, Mindset Ventures, and Vintage Investment Partners; over $100M raised in the prior 12 months.
Founded
2018
Headquarters
New York, USA and Tel Aviv, Israel
Stage
Series C
Related companies
Entera
AI-powered platform that helps institutional single-family investors find, evaluate, and transact on properties at scale across 32 US markets.
Discovery algorithms scan single-family inventory every 15 minutes across 100+ on and off-market sources, then match each property to an investor's buy box and predict its financial returns.
HouseCanary
AI-powered residential real estate analytics platform delivering automated valuations, property data, and market forecasting across 136M+ properties.
HouseCanary runs machine-learning valuation models, image recognition, and a generative AI assistant across data on more than 136 million US residential properties.
CityBldr
CityBldr is a real estate intelligence platform that uses machine learning to locate, value, and analyze underutilized properties and development sites. It serves developers, brokers, investors, and asset managers across US and Canadian markets.
CityBldr applies machine learning to public and proprietary real estate data to surface, value, and analyze underutilized development sites and parcel assemblages across roughly 155 US and Canadian cities.