DevRev adds shared memory to enterprise Computer AI

DevRev has released a new version of its Computer AI product for enterprise teams, adding shared memory, multiplayer collaboration, a desktop app, and tools for building AI agents.

The update targets a common complaint about workplace AI: systems can generate answers quickly but often lack business context, forcing staff to check, correct, and rework the output. DevRev linked that issue to broader concerns about whether AI is improving productivity in practice.

Data cited from the Upwork Research Institute found that 96% of C-suite leaders expect AI to raise productivity, while 77% of employees said it had increased their workload. DevRev argued that prioritising speed and token volume over context and reliability has created more noise for knowledge workers.

The latest version of Computer is built around what DevRev calls shared memory. The system is designed to retain context at the individual, team, and organisational levels, so sessions build on previous work instead of starting from scratch each time.

At the individual level, the software retains how a person works and can continue from earlier sessions. At the team level, workflows and agents created by one user can be reused by colleagues. At the organisational level, institutional knowledge stays in the system, including information tied to specific employees who may later leave the business.

Collaboration tools

One addition is Multiplayer AI, which lets teams work in a shared live session. That allows colleagues to review, question, and amend the same chain of reasoning in real time rather than use AI tools in isolation.

DevRev also pointed to research from KPMG and the University of Melbourne, which found that 57% of employees admitted using AI in non-transparent ways, including not disclosing when they had used AI tools to complete work. Shared sessions are intended to shift AI use from an individual activity to a visible team process.

Another feature is a desktop application that produces business documents and other outputs from an in-app canvas. Users can generate items such as slide decks, quarterly business review reports, dashboards, knowledge base articles, and multi-step workflows using underlying business data.

Those outputs can be exported in formats including PPT, PDF, HTML, and DOCX. Workflows and materials created in the canvas can also be saved and reused by individuals, teams, or across the wider business.

Agent controls

The release also includes Agent Studio, a tool for creating, testing, and deploying AI agents across connected systems. DevRev said agents operate under the permissions of the individual user rather than through a shared account, and that each action can be traced, audited, and reversed.

The broader product includes search and analytics functions, along with connectors to workplace software such as Gmail, Outlook, Slack, Notion, Google Drive, Jira, Microsoft OneDrive, and SharePoint. It is available through web, mobile, and desktop interfaces, and can also be accessed within DevRev's own system-of-record applications.

Dheeraj Pandey, Co-founder and Chief Executive Officer of DevRev, said the launch responds to growing strain on workers using AI systems that produce output without enough context.

"Every AI company is selling speed, which is fueling the next enterprise crisis. This will be the latest form of workplace burnout if we don't do something about it," Pandey said.

"Speed without the right context is just faster noise, noise that overloads humans in the loop, and eventually breaks them. Computer is built on a different philosophy: work softer. Give AI the enterprise memory and shared context it needs to perform reliably, and then let it take action. Only then will your people have the confidence that the AI they use is performing the tasks they need it to, more accurately and with less handholding," he said.

Customer use

More than 250 organisations are using Computer in production, with over 1,000 users onboarded since its launch in September 2025, according to DevRev. Customers named by the company include BILL, HDFC Bank, and FAME, across sectors including financial services, aviation, retail, and technology.

Some customers, including Pebl and Uniphore, are resolving 85% of support tickets without human involvement, DevRev said. BILL has recorded about USD $5 million in operational savings, while an unnamed airline in India moved from kickoff to production in 14 days, according to the company.

DevRev also said a retail loyalty customer is saving USD $1.2 million a year, with sales staff reclaiming six hours a week and reporting a 30% productivity increase. FAME is saving users more than 10 hours a week, resolving tickets about 40% faster, and cutting the time for some workflows by as much as 75%, the company said.

Computer is now generally available across the company's supported platforms, with usage-based pricing tied to adoption rather than headcount.