Convr: IDP Software Vendor

AI underwriting workbench for commercial P&C insurance with agentic workflow agents, a generative AI assistant, and 97% document extraction accuracy.

Overview

Founded in 2016 by underwriters Harish Neelamana and Kuldeep Malik, Convr builds AI-powered tools for commercial property and casualty (P&C) insurance underwriting. The company raised $15-18.2M from investors including Altos Ventures, NFP Ventures, Palm Drive Capital, and Stage 2 Capital.

Three product moves in early 2026 define where Convr is heading. In January, the company launched agentic AI workflow agents for autonomous underwriting decisions, shifting from rule-based automation to proactive AI that operates without waiting for commands. In March, Convr embedded a generative AI assistant directly into the workbench, built on a proprietary Context Engine combining a commercial insurance ontology, knowledge graph, and semantic layer. That same month, the company expanded its Data Catalog to surface nearly 3,000 external data sources with refresh frequency metadata, giving underwriters visibility into which data is current and reliable.

Enterprise validation has accumulated across multiple relationships. Zurich North America expanded their 2017 partnership across multiple business units after an 8-year relationship. Lucid Insurance Group achieved straight-through processing with complete automation from submission intake through BindHQ policy administration. In February 2026, Ohio Mutual Insurance Group (OMIG) deployed all three core workbench modules to replace what OMIG described as "legacy, siloed tools," a three-module deployment that signals a consolidation play rather than a single-feature pilot.

Convr positions itself as the "fourth core system" alongside policy, billing, and claims systems, targeting infrastructure-level adoption. The company self-reports a 70% reduction in submission-through-quote times; no methodology or third-party validation accompanies that figure.

A Convr survey of 200 commercial P&C decision-makers found 83% plan to adopt new AI tools but only 47% currently use technology for underwriting. That adoption gap is the market Convr is positioned to close.

How Convr processes documents

Convr's processing pipeline begins with Intake AI, a patented intelligent document automation layer that ingests, splits, classifies, and prioritizes data from structured and unstructured documents. Supported formats include ACORDs, loss runs, statements of values (SOVs), schedules, financial statements, and broker forms. The system achieves 97% extraction accuracy in approximately 30 seconds, compared to hours for manual processing. Output is delivered as JSON via a commercial P&C ontology engine developed since 2016, providing a unified data structure for downstream integration with policy systems and rating engines.

Extracted data feeds into Risk 360, which aggregates business insights from thousands of public, private, and customer-preferred data sources alongside the platform's 85+ million business record data lake. As Chris Healey from Zurich North America notes, this "goes far beyond data extraction" to improve "understanding of the true risk profile" through enriched exposure data, surfacing prioritized results in minutes rather than requiring manual research.

The March 2026 generative AI assistant adds a conversational layer on top of this pipeline. Built on Convr's proprietary Context Engine, which combines a commercial insurance ontology, knowledge graph, and semantic layer powering a multi-line schema, the assistant lets underwriters retrieve submission details, summarize risk data, and answer underwriting questions in real time. Critically, it can also take action: finalizing submissions, creating tasks, and updating records without leaving the workbench interface. Co-Founder and President Harish Neelamana described the capability as "similar to ChatGPT where we can ask any observations or questions about a submission and the result is an accurate account or summary of the results. Following a prompt, it can actually take an action and finalize the submission for the underwriter." Conversations are memorialized in the underwriting file for analysis and usage improvement.

Enriched data and conversational context then feed the agentic workflow layer, where autonomous agents handle referral, declination, financial analysis, and underwriting authority functions using pre-built templates that require no prompt engineering. Field-level confidence scoring enables autonomous decision-making while preserving human-in-the-loop quality control for exceptions. For a broader view of how autonomous agents are reshaping document workflows across industries, see the agentic document processing capability overview.

The full pipeline supports straight-through processing (STP): Lucid Insurance Group runs complete automation from submission intake through BindHQ policy administration with no manual handoffs, serving as the reference architecture for end-to-end deployment.

Use cases

Commercial P&C carriers: autonomous underwriting workflows

Insurance carriers deploy Convr's agentic AI agents to automate underwriting decisions across the submission lifecycle. The agents operate on pre-built templates covering four primary decision points: referral, declination, financial analysis, and underwriting authority, without requiring custom prompt engineering for each carrier's workflow. Zurich North America's expanded deployment across multiple business units, beginning with automobile lines, illustrates how carriers scale from a single line of business to enterprise-wide adoption.

Specialty commercial lines: replacing fragmented tool stacks

Ohio Mutual Insurance Group deployed all three core modules specifically to consolidate away from "legacy, siloed tools" into a centralized underwriting repository. The competitive frame here is not a single rival product but the fragmented point-solution stacks that commercial lines underwriters typically assemble. Gary Johnson, VP of Commercial Lines at OMIG, cited anticipated gains in "workflow, accuracy, and service" once the platform is operational; quantified benchmarks are not yet disclosed.

MGAs and insurtech carriers: straight-through processing

Lucid Insurance Group achieved complete STP from submission intake through BindHQ policy administration, full automation with no manual handoffs between systems. This deployment serves as Convr's reference architecture for carriers and MGAs targeting end-to-end automation rather than point-solution efficiency gains. Indico Data and SortSpoke address adjacent insurance submission automation use cases and offer a useful comparison point for evaluators assessing the competitive landscape in commercial P&C. Cytora takes a comparable approach to commercial insurance submission digitization and underwriting automation, making it another relevant reference for evaluators comparing purpose-built insurance IDP platforms.

Technical specifications

Resources

• Workbench
• Generative AI Assistant announcement
• Ohio Mutual Insurance Group partnership
• Data Catalog enhancement
• Agentic AI Workflow Agents launch
• Zurich North America expansion

Company information

Headquarters: Schaumburg, Illinois, United States

Founded: 2016

Founders: Harish Neelamana (Co-Founder and President), Kuldeep Malik

CEO: John Stammen

Head of Data and AI Underwriting Solutions: Eli O'Donohue (hired February 2026; previously Head of Data Strategy at PCMI, Director of Data Solutions at Planck, Senior Director of R&D at Carpe Data)

Funding: $15-18.2M (Altos Ventures, NFP Ventures, Palm Drive Capital, Stage 2 Capital)