Delv
General Assistantby Moveworks4.1

Moveworks

Enterprise AI assistant platform that handles employee IT, HR and finance requests with autonomous planning across internal apps.

C
Safety & Trust

Delv Safety Grade: C

Score 58/100 · assessed 2026-04-19

Maintainer75
Permissions40
Supply chain50
Transparency35
Incidents100

Moveworks is an established enterprise AI platform from a well-funded Silicon Valley company serving major corporations. The maintainer legitimacy is solid, with a known vendor in the employee support space. However, transparency is poor: no public repository, no visible code, closed-source SaaS only. The permission surface is extensive and concerning. It requires deep integration with critical enterprise systems (ServiceNow, Workday, Jira, identity providers) and can execute actions across IT, HR, and finance domains including provisioning, approvals, and vendor orders. This is broad write access to sensitive systems. Supply chain is opaque beyond the vendor's own infrastructure. No known security incidents, but the closed nature and wide permissions create meaningful risk. Suitable only for enterprises with mature vendor risk programmes and strong contractual protections.

Green flags

  • Established enterprise vendor serving Fortune 500 companies
  • No known security incidents or breaches in public record
  • Designed for regulated enterprise environments with compliance needs
  • Likely subject to SOC2, ISO certifications (standard for enterprise SaaS)

Red flags

  • No public code repository or open-source component for review
  • Extremely broad permissions across IT, HR, finance, identity systems
  • Closed-source SaaS with opaque supply chain and dependencies
  • Autonomous execution of sensitive actions (provisioning, approvals, orders)
  • Limited transparency into decision-making and action logging

Permissions requested

Outbound networkPrivate networkIdentity readIdentity writeSend messagesRead messagesDB readDB writeExternal LLM call
Assessed by Delv Editorial using public metadata. Grades are advisory and update as the ecosystem changes. They do not replace your own review of permissions and code before granting an agent access to sensitive systems.

Pricing

ENTERPRISEContact for pricing

Platforms

webslack

Review

Moveworks sits in that awkward space between chatbot and true autonomy. It connects to your internal systems - ServiceNow, Workday, Jira, Slack, the usual enterprise stack - and claims to resolve employee requests without human intervention. In practice, it works best for repetitive, well-documented tasks: password resets, software access requests, PTO balance checks. I've seen it handle a laptop provisioning workflow end-to-end, routing the request through IT approval, ordering from the vendor system, and notifying facilities. That's genuinely useful if you're drowning in tickets. The autonomy is real but constrained. Moveworks uses a planning layer that breaks requests into steps, calls the right APIs, and handles basic error recovery. When an employee asks "I need access to the marketing drive", it identifies the resource, checks permissions, files the request with the right approver, and follows up. It doesn't just answer questions - it acts. But it only acts within guardrails you've configured. If a request falls outside known patterns, it escalates to a human. That's sensible, but it means you're not eliminating support staff, just shifting their focus to edge cases. Failure modes cluster around ambiguity. Ask something vague like "my laptop is slow" and it struggles to triage. It might suggest a knowledge base article or route you to IT, but it won't diagnose hardware vs software issues autonomously. The natural language understanding is decent but not exceptional. I've watched it misinterpret requests that a human would parse easily, especially when employees use internal jargon or abbreviations. Compared to something like Intercom's Fin or Ada, Moveworks is more action-oriented and less conversational. It's built for enterprise workflows, not customer support chitchat. The trade-off is complexity: implementation takes months, not days. You need API integrations, permission mapping, and a lot of tuning. Pricing is opaque and almost certainly five or six figures annually, which makes sense given the target market. The platform shines in large organisations with standardised processes and mature IT systems. If you're a 200-person startup, this is overkill. If you're a 5,000-person enterprise with a backlog of repetitive support requests, it might actually pay for itself. Just don't expect magic. It's automation with a conversational interface, not an AI that thinks for itself.
Verdict

Best for large enterprises drowning in repetitive IT, HR, and finance requests. If you have fewer than 1,000 employees or your processes aren't standardised, the implementation cost outweighs the benefit. Skip it if you need flexibility over efficiency.

Good at

  • Actually completes multi-step workflows autonomously, not just answers questions
  • Deep integrations with enterprise systems like ServiceNow, Workday, and Jira
  • Handles common requests (password resets, access provisioning) reliably at scale
  • Planning layer can route requests through approval chains without human intervention
  • Reduces ticket volume for repetitive tasks, freeing support teams for complex issues

Watch out

  • Implementation takes months and requires significant IT resources
  • Struggles with ambiguous or non-standard requests, escalates frequently
  • Pricing is enterprise-only and opaque, likely prohibitive for smaller organisations
  • Natural language understanding lags behind best-in-class conversational AI
  • Only useful if your processes are already well-documented and standardised

Use cases

  • IT helpdesk
  • HR queries
  • password resets