DevRev
DevRev runs Computer, an enterprise AI platform built on a persistent, permission-aware shared memory. It searches live company data, resolves support and IT tickets, updates CRM records and lets any team build no-code agents.
What is DevRev?
DevRev is a Palo Alto company whose product, Computer, is sold as a single enterprise AI platform meant to serve every team rather than one department. Its stated differentiator is Native Shared Memory: a persistent, permission-aware memory that reasons over an organisation's live data instead of querying a periodically refreshed index. The layer underneath is Computer Memory, a patented knowledge graph described as a digital twin of the company's data, pulling CRM records, tickets, emails and Slack threads into one structure that grows richer with each session and decision.
Data reaches that graph through AirSync, a patented sync engine. The free tier ships one-click connectors for Slack, Notion, Jira Cloud, Google Drive, Google Calendar, Microsoft Teams, OneDrive and SharePoint; the paid tiers open a marketplace of more than a hundred connectors and snap-ins, from Salesforce and Zendesk to Okta, ServiceNow, Gong, Figma and BambooHR. Three companion applications sit on top: a Support App for customer service, a Build App for issues and roadmaps, and an Observe App for user behaviour, feedback and sentiment.
Computer is designed to act, not only to answer. It resolves tickets, updates CRM records and sends follow-ups, but every action is logged, traceable and reversible, and consequential ones such as refunds or external communications stop for human approval. Agents are proved before they ship, through sandbox testing, bulk evaluation against real datasets and side-by-side version comparison. Agent Studio lets teams assemble those agents without code, and a public API plus the SnapKit toolkit cover custom development. Sessions are multiplayer by design: teammates can be added to any conversation and co-create in a shared Canvas.
The company was founded in 2020 by Dheeraj Pandey, previously of Nutanix, and Manoj Agarwal. It reports more than 650 employees across eight offices and a valuation above 1.2 billion USD. Its published performance figures — 48% more accurate on 77% fewer tokens — come from Enterprise-Bench, a benchmark DevRev itself publishes, and should be read as a vendor claim.
What it does
- Search every system the organisation has connected and return cited answers drawn from live data
- Resolve level-1 and level-2 customer support tickets directly, on every channel
- Answer and close IT and service desk requests automatically, 24 hours a day
- Draft decks, prepare meetings, summarise pipeline opportunities and risks, and update the CRM
- Build, test, deploy, observe and optimise agents through Agent Studio, without writing code
- Unify data from Salesforce, Jira, Zendesk, Slack, GitHub and a hundred more connectors
- Track project progress, surface risks early and push status updates on behalf of ops teams
When to use DevRev / When not to
A quick filter to help you decide if DevRev is the right fit.
When to use DevRev
- Customer support teams working through large ticket backlogs across several channels at once
- Sales teams that want deck drafts, meeting prep, pipeline summaries and CRM updates handled for them
- IT and service desk managers automating level-1 and level-2 employee requests around the clock
- Large organisations that need org-wide search over live product, support and sales data rather than stale indexes
- Seed or Series A startups under five years old, who qualify for twelve months free and up to 10,000 USD in credits
When not to use DevRev
- Buyers who need a published price before they will talk to a salesperson: the two paid tiers only say "Contact us"
- Small teams content with the free Mini tier, since Agent Studio, the connector marketplace and conversational analytics all sit behind a quote
- Organisations that must document a specific hosting country or region before signing, as neither is published
- Companies that need a configurable opt-out from model training rather than a blanket vendor commitment
- Competitors of DevRev, whom section 4.1 of the terms of service explicitly bars from accessing the service
How to use DevRev
A typical end-to-end flow, from setup to results.
- Sign up directly for the free Mini tier at the DevRev application, or book a demo if you are aiming at Pro or Max
- For the paid tiers, take the 30-minute, no-commitment demo, which DevRev runs on your own systems
- Connect your first data sources with the one-click connectors: Slack, Notion, Jira Cloud, Google Drive and Calendar, Microsoft Teams, OneDrive and SharePoint
- Choose which folders and documents Computer may read, so that meeting notes, project docs and working files become part of its memory
- Ask questions across the connected systems and check the citations and the reasoning steps behind each answer
- On Pro, add marketplace connectors and configure them, then switch on conversational analytics for Text2SQL queries
- Build your first agent in Agent Studio without writing code, then test it in the sandbox and evaluate it against real datasets
- Set the approval boundaries: decide which routine actions run immediately and which high-stakes ones must wait for a human
- Deploy across surfaces — desktop app, mobile app and browser — with chats, memory and context kept in sync
- Train the team on DevRevU, consult the documentation, and use SnapKit and the public API when you need something bespoke
Pros & Cons
Pros
- An unusually dense compliance file: ISO/IEC 27001:2022, SOC 2 Type 2, HIPAA, CCPA and GDPR, with penetration test and vulnerability reports available on request
- A written commitment that customer data is never used to train generalised AI models
- A published, downloadable data processing agreement supplied at sign-up, with standard contractual clauses and the UK addendum
- A named Article 27 representative inside the EU, an appointed data protection officer and a published sub-processor list
- A permanent free tier that already includes org-wide search and one-click connectors
- A broad, individually documented connector ecosystem covering roughly fifty named tools
- Operational guardrails that matter in practice: approval gates, logging, reversibility and pre-production agent evaluation
Cons
- No public price for either paid tier, so no budget can be estimated without a sales conversation
- The credit model is described but its rates are never published
- The decisive capabilities — Agent Studio, the marketplace, Text2SQL and the Computer+ apps — all sit outside the free tier
- Neither a hosting country nor a hosting region is disclosed for customer data
- No configurable training opt-out is documented; the no-training commitment is unilateral and global
- No interface or processing language other than English is documented anywhere on the site
- The headline performance gains come from Enterprise-Bench, a benchmark published by DevRev itself
Pricing & Plans
A permanent free plan is available. The Mini tier is offered at no cost and can be opened by self-service sign-up. The two paid tiers, Pro and Max, both display "Contact us" instead of a figure, so no lowest paid price point and no currency are published. DevRev describes its commercial model as consumption-based, with customers buying only the credits they need and every dollar spent converted into credits; the credit rate itself is not disclosed. No time-limited free trial is announced anywhere on the site: the free Mini tier serves as the entry point.
- conversational interface on desktop
- mobile and browser
- accurate search across the organisation's data
- basic task automation
- one-click connectors (Slack
- Notion
- Jira Cloud
- Google Drive and Calendar
- everything in Mini
- plus the marketplace of 100+ connectors and snap-ins
- conversational analytics with Text2SQL
- no-code Agent Studio
- Silver-level support
- discounted credits through subscription models
- and the Support App and Observe App add-ons
- everything in Mini and Pro
- plus an embedded team from the Applied AI department
- conversational analytics fine-tuned to the customer
- enterprise-grade security including SCIM user provisioning
- near real-time AirSync
- Gold-level support
- and the Max Support App and Max Observe App add-ons
Data, GDPR & hosting
A consolidated view of how DevRev handles your data.
GDPR overview
Implementation is concrete rather than declarative. A data protection officer is appointed and reachable at dpo@devrev.ai. An Article 27 EEA representative is named in full: DevRev, razvoj programske opreme, d.o.o., Ljubljana, Slovenia, with its registration and VAT numbers. The controller is identified as DEVREV, inc., Palo Alto, company registration 4656646. A data processing agreement drafted against Article 28 is downloadable and supplied at sign-up, carrying the EU standard contractual clauses and the UK International Data Transfer Addendum; transfers to the United States rely on the EU-U.S. Data Privacy Framework and its UK extension. Access, erasure, objection and consent-withdrawal rights are documented with a 45-day response commitment, the Slovenian Information Commissioner is named as supervisory authority, and data on people under 16 is not knowingly processed. The policy has been in force since 21 May 2026.
Who owns the data?
Section 7.1 of the terms of service leaves ownership of Customer Data entirely with the customer: everything you, your employees or your end users upload stays yours, along with sole responsibility for its accuracy, legality, correction and deletion. Section 7.3 reserves to DevRev all rights in the service, the software and the transactional and performance data it generates. The data processing agreement casts DevRev as processor under Article 28 of the GDPR and each customer as controller. DevRev says it keeps human oversight of its AI technologies, but section 10.4 places the duty of reviewing and validating any output before business use on the customer alone.
Reuse rights
Because customers keep title to their own data, they may use, export, correct or delete it without asking DevRev for permission. What DevRev itself may do is narrower than usual: the privacy policy states in plain terms that user and customer data is not used to develop, improve or train generalised AI or machine-learning models, and version 2.1 of the policy extended that clarification to data reached through Google Workspace APIs. Sub-processors operate under data processing agreements or standard contractual clauses and may not repurpose the data; the list of them is published on the site. Marketing email runs on consent that can be withdrawn at any time. Product usage data is analysed to improve the service, and anonymised or aggregated analytics may be kept for longer statistical work.
Data retention & training
Hosting summary
DevRev publishes no hosting country and no hosting region for customer data, which is a notable gap for a platform sold to regulated enterprises. What the documents do establish is jurisdiction: DevRev describes itself as a US-based organisation, and transfers from the EEA and the UK to the United States rely on the EU-U.S. Data Privacy Framework and its UK extension, certified with the Department of Commerce. Outside that framework, the company uses the European Commission's standard contractual clauses and the UK International Data Transfer Addendum. Many sub-processors are registered in the United States or other third countries, and the list is published. The trust centre records AWS Partner status alongside ISO/IEC 27001:2022, SOC 2 Type 2, HIPAA, CCPA and GDPR compliance, with penetration test and vulnerability reports available on request. One location is stated explicitly, but it concerns an internal tool: meeting transcripts and summaries produced by Granola are stored on servers in the United States. The funding announcement refers to a multi-region platform without naming a region.
Things to keep in mind
Risks and trade-offs to weigh before adopting DevRev.
- Pricing opacity: with both paid tiers quote-only and the credit rate unpublished, costs are hard to forecast and easy to underestimate before the first invoice
- Data location is undisclosed, which can block procurement in regulated sectors or under residency requirements
- The no-training commitment is a vendor pledge, not a switch you control; if it ever changes, you have no configured setting to fall back on
- A memory that remembers everything across the whole organisation raises real permission questions: an over-broad connector can surface documents someone assumed were private
- Automated actions on live systems — CRM updates, ticket resolutions, external messages — go wrong silently unless approval boundaries are set deliberately from the start
- Habituation: when answers always arrive pre-summarised and pre-cited, teams stop opening the underlying record, and errors travel further before anyone notices
- Vendor-published benchmark figures should be re-tested on your own data rather than accepted as measured performance
Setup & Integrations
Technical difficulty
Low to start, higher to scale. Opening the free Mini tier takes a self-service sign-up, and the first data sources connect in one click: Slack, Notion, Jira Cloud, Google Drive and Calendar, Microsoft Teams, OneDrive and SharePoint. Agents are assembled in Agent Studio without code, and search surfaces can be embedded in a website in a handful of lines. Beyond that, bespoke connectors and snap-ins need SnapKit and a developer, SCIM provisioning arrives only on the Max tier, and anything past the free plan starts with a sales demo. Free DevRevU training and public documentation soften the curve.
Deployment
Integrations
Behind DevRev
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement DevRev.
Frequently asked questions
How much does DevRev cost?
Is there a permanent free plan or a free trial?
Is my company data used to train AI models?
What security certifications does DevRev hold?
Is a data processing agreement available?
Where is my data hosted?
Does DevRev offer an API?
Which tools does it integrate with?
Is there a mobile app?
Is there an offer for startups?
Should you pick DevRev?
DevRev makes a coherent bet: rather than another assistant bolted onto one department, a single platform whose shared memory reasons over the company's live data, so that a support engineer and a sales rep asking the same question each get the answer their role needs. The plumbing behind that claim is real and documented — a patented knowledge graph, a sync engine, roughly fifty named connectors, no-code agent building, and guardrails that are genuinely operational: approval gates on high-stakes actions, full logging, one-click rollback and sandbox evaluation before an agent reaches customers.
Its compliance file is among the more serious you will find at this size. ISO/IEC 27001:2022 and SOC 2 Type 2, a downloadable data processing agreement handed over at sign-up, a named Article 27 representative in Slovenia, an appointed data protection officer, a published sub-processor list, and a written commitment never to train generalised models on customer data. For a buyer whose procurement team asks hard questions, that is a short due-diligence path.
The gaps are just as clear. Neither paid tier shows a price, the credit rate is unpublished, and no hosting country or region is disclosed — three answers most enterprise buyers will need before signing. The no-training pledge is unilateral rather than configurable. And the headline gains, 48% more accurate on 77% fewer tokens, come from a benchmark DevRev publishes itself, which makes them a claim to test rather than a measurement to trust.
The sensible route is the one the vendor already lays out: open the free Mini tier, connect a few real sources, and judge the shared memory on your own data before the pricing conversation begins.
- Choosing a selection results in a full page refresh.
- Opens in a new window.