Rohan Verma
Rohan Verma is an individual creator and software engineer who built SurfSense, an open-source, privacy-focused research and knowledge agent. SurfSense helps people research, transform, and edit their own documents and, in its newer direction, gives AI agents structured access to live web data while keeping self-hosting and local use available.
At a Glance
- Law firms and legal teams
- Accountants and auditors
- Consultants and analysts
- Privacy-conscious individuals and developers
- +2 more
AI Tools by Rohan Verma
(1)SurfSense
Open Source Local Document AI Agent
Discussions
No discussions yet
Be the first to start a discussion about Rohan Verma
Latest News
SurfSense desktop release 2.2.0 shipped with expanded file-agent, editing, Office/PDF, model and Agentic-mode capabilities.
SurfSense announced the hosted cloud service is moving to a local app; cloud data is export-only and scheduled for deletion on October 18, 2026.
Reports, resumes and other deliverables moved to the artifacts system, while document export was removed from the editor.
SurfSense announced its next chapter as an open-web research platform for AI agents, adding platform-native connectors, REST access, MCP, agent infrastructure and automations.
Products & Services
Rohan Verma's main open-source project, evolving from a Chrome-extension personal knowledge base into a self-hosted/web knowledge workspace and air-gapped desktop app. It researches documents with cited answers, creates and edits files, generates artifacts such as decks, reports, briefings and podcasts, supports local or remote models, and is expanding into structured open-web connectors and agent tooling.
A free local application for Windows, macOS and Linux that runs without an account, keeps the index on the user's disk, supports local or remote models, and imports workspaces exported from the hosted service.
Structured REST/MCP access for live data from Reddit, YouTube, Google Maps, Google Search and the open web, plus agent retries, structured output, credit metering, external MCP connectors and automations.
Market Position
SurfSense positions itself as an open-source, privacy-first and air-gapped alternative to Google NotebookLM and other hosted research/knowledge assistants. Its differentiation is that documents can stay on the user's machine, models can be local or provider-independent, files can be edited rather than merely queried, and the newer platform direction exposes live web research primitives through REST and MCP for agents. Its open-source/self-hosted posture also contrasts with closed hosted tools such as NotebookLM and Perplexity.
Leadership
Founders
Rohan Verma
Creator and founder/CTO of SurfSense. His LinkedIn profile identifies him as a software engineer with 4+ years of experience building scalable applications and enterprise solutions; reporting from Xseed says he began SurfSense during his master's program in summer 2024 to combine research from files and the web.
Executive Team
Rohan Verma
Founder and CTO of SurfSense; creator
Software engineer with 4+ years of experience building scalable applications and enterprise solutions; Northeastern University affiliation is shown in LinkedIn search results. He started SurfSense during his master's program in 2024.
Founding Story
SurfSense began as Rohan Verma's personal project during his master's in summer 2024. He needed an agent that could answer assignments using both file-based content and online research, built the smallest useful version, posted it on Reddit, and iterated based on usage. The initial public positioning was a privacy-preserving personal knowledge base for saving, searching, and interacting with web content.
Business Model
Revenue Model
The project is free and open source for the app/self-hosted use. The hosted/cloud direction uses pay-as-you-go billing: $5 of free credit to start, connectors billed per returned item, crawls per fetched page, and models at provider cost; enterprise deployments are sold by contract. The local desktop app keeps its core features free, while paid licensing is for scraper plugins and priority support.
Pricing Tiers
Free forever, no account, no trial clock and no usage cap for the app and updates.
A 30-day license is delivered by email and unlocks scraper plugins and priority support; the pricing page says it covers nine platform connectors plus the web crawler.
The July 5 changelog says cloud is pay as you go; a SurfSense blog says the no-account /free chat provides 500,000 free tokens across available models.
Private deployment with no egress, mirrored licensing/plugin endpoints, centrally set model and network policy, local outbound-call audit logs, pinned installers, internal connectors and named support.
Target Markets
- Law firms and legal teams
- Accountants and auditors
- Consultants and analysts
- Privacy-conscious individuals and developers
- Organizations with confidential, regulated or air-gapped data
- AI-agent builders and teams needing live web data
- Private research over documents that cannot be uploaded
- Legal, accounting, audit, consulting and analyst workflows involving confidential documents
- Personal knowledge management and cited question answering
- Creating and editing business documents, presentations, spreadsheets and PDFs
- Offline or air-gapped AI deployments
- AI-agent web research, competitive intelligence, review and ranking monitoring