Professional services firm
PwC (PricewaterhouseCoopers LLP, US)
PwC US offers agent OS, a platform for orchestrating AI agents, and AI-enabled finance managed services; PricewaterhouseCoopers LLP audits public companies.
Function
- Ownership
- Not stated
Not stated in the sources read. PwC says its network member firms are each a separate legal entity. Neither PricewaterhouseCoopers LLP entity on EDGAR files Forms 10-K, 10-Q or 20-F.
- Headquarters
- United StatesSourceInspection Data for U.S. Global Network Firms
- SEC reporting
- NoSourceEDGAR company searches and submission records for the audit and advisory firms (Deloitte, PricewaterhouseCoopers, Ernst & Young, KPMG, Grant Thornton, BDO USA, RSM US, Protiviti)
- Audit firm
- YesSourceInspection Data for U.S. Global Network Firms
- Subprocessor list
- Not stated
- Contract terms
- Not stated
Offering · Accelerator
PwC's agent OS
PwC describes agent OS as a unified platform to orchestrate multiple AI agents, including agents from different vendors, into workflows with oversight built in, and lists transforming the finance function among its uses. SourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
It lists Anthropic, AWS, Google Cloud, Microsoft Azure, OpenAI, Oracle, SAP, Workday and others as platforms and enterprise systems agent OS works across. SourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
Its update log says agent OS supports OpenAI's GPT-5 and that Anthropic's Claude Haiku 4.5 is available within agent OS via Amazon Bedrock. SourceAgent OS recent enhancements (dated update log)
PwC says its orchestration enables agents to act within workflows, and describes drag-and-drop building, a three-tier role model, Model Context Protocol access to enterprise tools, and session tracking with human-in-the-loop feedback. SourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
Where it fits
CoE analysis An orchestration layer rather than a finance application; it matters to finance where agents from several vendors would act across ERP and planning systems.
PwC's own figures on speed and effort reduction are vendor-stated. SourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
What it does
- Offering type
- AcceleratorSourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
CoE classification: a firm-owned platform from an advisory practice. PwC does not say whether agent OS is licensed, hosted or used only within engagements.
- Workflow areas
- Not stated
PwC lists transforming the finance function among many uses without naming finance workflows.
- Customer size
- Not stated
AI
- Kind of AI
- Agents and automated actionsSourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
- Suggests or acts
- ActsSourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
- Availability
- Not stated
PwC says agent OS is constantly evolving; no availability label.
- Models named
- SourceAgent OS recent enhancements (dated update log)
- Claude Haiku 4.5 (Anthropic), available within the platform
- GPT-5 (OpenAI), supported within the platform
Control
- Who controls it
- Individual userSourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
Non-programmers can drag and drop agents into workflows; a User, Superuser and Admin role model governs access. Which client function holds the Admin role is not stated.
Commercial
- Pricing model
- Not stated
- Public price
- Not stated
Evidence
- Independent outcomes
- None found in the sources read.
- Sources read
- Vendor-stated 6
- Independent market evidence 5
- Independently verified outcomes None read
- Pricing None read
- Known limitations None read
- Coverage
- Directory record
- Last verified
Delivery and lock-in
- Delivery model
- Not stated
Not stated in the sources read. PwC calls agent OS a unified platform and says it can run within AWS and directly in Oracle Cloud Infrastructure environments; whether clients subscribe to, license or keep it is not stated.
- Run service offered
- Not stated
- Team qualifications published
- Names Matt Labovich, Advisory AI Leader, PwC US, as the contact. No delivery team, team size or credential criteria are stated.SourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
- IP and post-engagement license
- Not stated
Not stated in the sources read. PwC calls its orchestration technology patent-pending and says agents and workflows built in agent OS can be reused, but states no license terms for clients during or after an engagement.
- Savings claims (firm-stated)
- SourcesPwC's agent OS: Integrate and govern enterprise AI agents (service page)Agent OS recent enhancements (dated update log)
Stated by the firm, not verified
Automating entire workflows with agent OS can reduce manual effort by up to 95%.
Not stated; no client, baseline, period or cost basis is given. PwC also says results can come in as little as 30 days if a client's tech stack is ready.
Efficiency gains such as reducing call-center times by 25%.
Not stated; client, baseline and period not given. Not specific to finance.
Shortening compliance review workloads by as much as 94%.
Not stated; client, baseline and period not given.
Build, integrate and scale AI-driven processes up to ten times faster than traditional methods.
A speed claim rather than a cost figure; basis not stated.
- Alliance partners
- SourceAgent OS recent enhancements (dated update log)
- OpenAI
PwC describes its participation as an OpenAI Services Partner. It also lists Anthropic, AWS, GitHub, Google Cloud, Microsoft Azure, OpenAI, Oracle, Salesforce, SAP, Workday, CrewAI and LangGraph as platforms agent OS works across (C-4226), and describes AWS and Oracle Cloud deployment support (C-5013), without calling them partners on the pages read.
Security and control
- Single sign-on
- Not covered in this record
- User provisioning
- Not covered in this record
- Role-based access
- Not covered in this record
- Administrator controls for AI
- Not covered in this record
- Audit logs
- Not covered in this record
- Certifications
- Not covered in this record
- Data retention
- Not covered in this record
- Data residency
- Not covered in this record
- Training on customer data
- Not covered in this record
AI features
As the vendor or firm describes them. Notes in small print record what the sources leave open.
Agent orchestration
Connects agents, including other vendors' agents, into workflows and enterprise-wide processes; PwC says its orchestration enables agents to act within workflows based on real-time data and natural language.
- Kind
- Agents and automated actions
- Suggests or acts
- Acts
- Who controls it
- Individual user
- Availability
- Not stated
Control: Drag-and-drop building for non-programmers; SDKs for engineers.
SourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)Governance dashboard
Session tracking, execution history and human-in-the-loop feedback in a centralized dashboard; sandboxed testing; MCP-based access to tools and data.
- Kind
- Agents and automated actions
- Suggests or acts
- Not stated
- Who controls it
- Individual user
- Availability
- Not stated
Control: Role-based access with User, Superuser and Admin roles.
SourcePwC's agent OS: Integrate and govern enterprise AI agents (service page)
Questions to ask
For this offering
- If PwC audits our financial statements, has our audit committee pre-approved this service, and how does PwC assess it against the SEC non-audit service categories (bookkeeping, financial information systems design and implementation, internal audit outsourcing, management functions) and, for ICFR-related work, PCAOB Rule 3525? Ask our audit committee and counsel to confirm.
- If PwC builds or operates agents on agent OS for us, which finance systems could they write to, and who approves their actions?
- Is agent OS licensed to us as software, hosted by PwC, or available only within an engagement, and on what fee basis?
- Which model providers would process our data, under whose contracts, and with what retention and training terms?
- What does the governance dashboard log, how long is it kept, and can our internal and external auditors access it?
From the diligence question set
- DQ-31 Who, by name and role, will design, build, test and run the AI in our finance process, and what AI has each of those people personally put into production?
- DQ-32 How does your firm decide that someone is qualified to build or govern AI for financial reporting, and who outside the delivery team reviews model risk before anything goes live?
- DQ-33 When the engagement ends, what will we own and be able to run ourselves, and what will work only with your firm? On what terms are your accelerators licensed afterward, and what does it cost to leave a run service?
- DQ-34 What will this cost in years two to five, including run fees, platform licenses, AI usage fees and change requests, and how does that compare with the savings case once the internal oversight roles we keep are counted?
- DQ-35 Do you receive fees, rebates, resale margins or other incentives from the software you are recommending, and do you, or will you, audit our financial statements?
- DQ-36 What results from clients like us can you show that someone other than your firm has verified, and by what method?
Offering · Managed service
PwC finance and accounting managed services
PwC describes finance managed services that range from AI-powered planning and analysis hubs to end-to-end controllership and treasury support, with AI embedded in operations and US-led teams backed by global centers. SourceFinance and accounting managed services (service page)
Its controllership services cover general accounting, period-end close, reconciliations, financial reporting and compliance, with embedded controls, workflow automation and AI-enabled monitoring. SourceFinance and accounting managed services (service page)
PwC calls the services outcome-based, aligned to business impact rather than hours worked. SourceFinance and accounting managed services (service page)
Where it fits
Controllership, close and planning work handed to PwC under a managed model, with AI in the delivery. SourceFinance and accounting managed services (service page)
CoE analysis The outcome-based commercial model and the auditor's position both need clarity before contracting.
What it does
- Offering type
- Managed serviceSourceFinance and accounting managed services (service page)
- Workflow areas
- SourceFinance and accounting managed services (service page)
- Close
- Reconciliation
- Reporting and disclosure
- Planning and FP&A
- Treasury and cash
- Receivables and collections
- Revenue and billing
- Customer size
- Not stated
AI
- Kind of AI
- Not stated
Not classified. PwC describes AI embedded in operations, AI-enabled monitoring in controllership, AI-powered forecasting in planning hubs, and continuous improvement pods that use AI, without naming the technique or saying whether any of it acts in client systems.
- Suggests or acts
- Not stated
- Availability
- Not stated
No availability statement.
- Models named
- Not stated
Control
- Who controls it
- Professional services firmSourceFinance and accounting managed services (service page)
Delivered by PwC teams; client approval points are not stated.
Commercial
- Pricing model
- Not stated
- Public price
- Not stated
Evidence
- Independent outcomes
- None found in the sources read.
- Sources read
- Vendor-stated 3
- Independent market evidence 7
- Independently verified outcomes None read
- Pricing 1
- Known limitations None read
- Coverage
- Directory record
- Last verified
Delivery and lock-in
- Delivery model
- Service you rentSourceFinance and accounting managed services (service page)
PwC describes outcome-based finance managed services, aligned to business impact rather than hours worked.
- Run service offered
- YesSourceFinance and accounting managed services (service page)
- Team qualifications published
- Describes US-led delivery teams supported by scalable global centers for cost efficiency and 24/7 coverage, and names Bob Woods, Partner, Finance Managed Services, PwC US, as the contact. No team size or credential criteria are stated.SourceFinance and accounting managed services (service page)
- IP and post-engagement license
- Not stated
Not stated in the sources read.
- Savings claims (firm-stated)
- Not stated
Not stated in the sources read. The page says continuous improvement pods help uncover cost savings, without a figure.
- Alliance partners
- Not stated
Not stated in the sources read.
Security and control
- Single sign-on
- Not covered in this record
- User provisioning
- Not covered in this record
- Role-based access
- Not covered in this record
- Administrator controls for AI
- Not covered in this record
- Audit logs
- Not covered in this record
- Certifications
- Not covered in this record
- Data retention
- Not covered in this record
- Data residency
- Not covered in this record
- Training on customer data
- Not covered in this record
AI features
As the vendor or firm describes them. Notes in small print record what the sources leave open.
Not stated
Questions to ask
For this offering
- If PwC audits our financial statements, has our audit committee pre-approved this service, and how does PwC assess it against the SEC non-audit service categories (bookkeeping, financial information systems design and implementation, internal audit outsourcing, management functions) and, for ICFR-related work, PCAOB Rule 3525? Ask our audit committee and counsel to confirm.
- How are outcome-based fees defined and measured, and how does PwC assess that structure where it is also the auditor?
- Which close, reconciliation and reporting steps would PwC staff or AI perform, and which approvals stay with us?
- What does the AI-enabled monitoring do, and is it covered by a SOC 1 report on the managed service?
- Which systems and data would PwC access, and from which delivery centers?
From the diligence question set
- DQ-31 Who, by name and role, will design, build, test and run the AI in our finance process, and what AI has each of those people personally put into production?
- DQ-32 How does your firm decide that someone is qualified to build or govern AI for financial reporting, and who outside the delivery team reviews model risk before anything goes live?
- DQ-33 When the engagement ends, what will we own and be able to run ourselves, and what will work only with your firm? On what terms are your accelerators licensed afterward, and what does it cost to leave a run service?
- DQ-34 What will this cost in years two to five, including run fees, platform licenses, AI usage fees and change requests, and how does that compare with the savings case once the internal oversight roles we keep are counted?
- DQ-35 Do you receive fees, rebates, resale margins or other incentives from the software you are recommending, and do you, or will you, audit our financial statements?