Pecan AI

Pecan AI develops a predictive analytics and predictive-modeling platform aimed at business intelligence analysts and non-technical business users rather than data scientists. The product, positioned as a "predictive AI agent" (earlier marketed as a low-code predictive modeling platform and, from 2025, a predictive "co-pilot"), connects to a customer's existing data warehouses, databases, CRMs and files through built-in connectors, runs an automated ETL and data-preparation step, generates SQL and engineered features, then trains, validates and evaluates models with data-science guardrails. Model performance is explained in business terms and natural language, and predictions are scheduled and pushed back into the customer's CRM, data warehouse or other operational systems rather than a separate dashboard. Users can also query predictions conversationally to generate insights and visualizations [0][1][2][4].

Supported use cases include customer churn prediction, customer lifetime value modeling, lead scoring, demand and inventory forecasting, upsell and cross-sell targeting, customer winback, fraud and chargeback prevention, and campaign ROAS prediction. Listed integrations span Snowflake, Databricks, Google BigQuery, Amazon Redshift, AWS S3, GCS/Parquet, Clickhouse, MySQL, PostgreSQL, Microsoft SQL, Oracle, IBM Db2, Salesforce, HubSpot, Firebase, and mobile attribution tools including Adjust, AppsFlyer and Singular [0][1][2]. The company also markets a related demand-forecasting offering, DemandForecast.ai [4]. Pecan states it maintains encryption and internal compartmentalization, endpoint and network protection measures for customer data [2].

Pecan was founded in 2018 by Zohar Bronfman (CEO) and Noam Brezis, who met during PhD studies in computational neuroscience at Tel Aviv University and are both alumni of IDF Unit 8200. The company operates from New York and Tel Aviv [5][6]. In February 2025 it appointed Trevor Healy as Executive Chairman and Andy Walter, a former Procter & Gamble CIO, to its advisory board [4].

Founding story

Pecan was founded in 2018 by Zohar Bronfman and Noam Brezis, who met while pursuing PhDs in computational neuroscience at Tel Aviv University; both are alumni of the Israeli military's Unit 8200. Bronfman serves as CEO [5][6].

Business model

Pecan sells its predictive analytics platform as software to companies, with pricing and demo-request paths on its website and a listing as a seller on AWS Marketplace for SaaS procurement [0][1][3]. The value proposition is replacing in-house data science and data engineering capacity: customers use the platform to build production models themselves, with predictions delivered into existing systems [3][5].

Software subscription/SaaS sold to midmarket and enterprise customers; the company has reported annual recurring revenue growth without disclosing absolute figures [5][6].

Traction

Pecan reported more than tripling annual recurring revenue in the year preceding its February 2022 Series C, with adoption among midmarket and enterprise customers, and headcount growth from 30 to about 100 employees over roughly nine months [5][6]. Named customers and references include Johnson & Johnson, which reported improved supply-chain forecasting accuracy and rapid model deployment, and Little Spoon, a direct-to-consumer food delivery service that used Pecan for lifetime-value and order-likelihood prediction to refine marketing spend and retention [4][5]. The company's site cites aggregate customer outcomes of roughly 12% average reduction in customer churn, 15% improvement in marketing ROAS, 25% reduction in inventory costs, and 90% of predictions delivered without data science support [0][1].

Latest developments

In February 2025 Pecan released a co-pilot for predictive modeling that uses conversational AI to let BI analysts refine data, explore trends and train models independently, announced alongside board additions Trevor Healy (Executive Chairman) and Andy Walter (advisory board) [4]. The company subsequently launched DemandForecast.ai, a GenAI-based supply-chain forecasting product [4], and now markets its core offering as a predictive AI agent that automates data prep, modeling and validation and delivers predictions into customers' existing tools [0][1][2].