AI agents are no longer a future talking point for revenue teams – they are shipping now, and the choices you make in the next six months will shape how your RevOps stack operates for years. Two significant releases have landed within days of each other: OpenAI’s ChatGPT Work, unveiled on July 9, 2026, and AWS’s agentic ERP automation layer, which brings deny-by-default security rules and a dedicated identity to AI-driven finance and operations workflows. Both tools target workplace automation, but they take very different approaches – and the right choice depends heavily on where your revenue operation’s biggest bottlenecks sit.
What Each Tool Actually Does
ChatGPT Work is OpenAI’s latest push into enterprise automation. Powered by the company’s most advanced models, it functions as a cross-application agent – meaning it can move between files, tools, and workflows inside a business environment rather than simply answering questions. Think of it as a generalist operator: it can draft follow-up emails after a deal stage change, pull context from a CRM record, summarise a prospect’s recent activity across documents, and hand a rep a ready-to-use call brief. For sales and GTM teams, the promise is a reduction in the low-value administrative work that eats into selling time.
AWS’s ERP agent is a more targeted instrument. Its headline use case is accounts receivable automation – specifically, matching incoming bank payments to open invoices without human intervention. For any organisation where days sales outstanding (DSO) is a meaningful metric, that matters. AWS has added deny-by-default permission rules and a separate identity for the agent, which signals that this is built for environments where audit trails and access controls are non-negotiable. The scope is narrower than ChatGPT Work, but the depth inside that scope is considerably greater.
Features and Use Cases for Revenue Teams
When you map each tool against common sales pipeline and finance workflows, the strengths become clearer.
ChatGPT Work is well suited to:
- Automating CRM data entry after calls, meetings, or email threads
- Generating personalised outreach sequences grounded in account context
- Summarising deal history for reps who are picking up accounts mid-sales cycle
- Drafting internal deal reviews aligned to qualification frameworks like MEDDIC
- Pulling together territory or segment reports without analyst involvement
AWS’s ERP agent is well suited to:
- Automated payment-to-invoice matching at scale
- Reducing blocked invoice queues that delay revenue recognition
- Supporting finance teams in keeping Monthly Recurring Revenue (MRR) reporting accurate and timely
- Providing auditable, permission-controlled AI actions in regulated industries
- Freeing AR staff from manual reconciliation so they can focus on exception handling
“Accounts receivable teams at large companies spend hours each day matching incoming bank payments to invoices by hand. When those payments sit unmatched for days, cash flow suffers and days sales outstanding climbs.” – AWS, via Help Net Security
The practical implication for RevOps leaders is that these tools are not really competitors. ChatGPT Work operates upstream in the revenue process – helping teams find, engage, and close customers. AWS’s ERP agent operates downstream – ensuring the revenue that has been closed is collected, recognised, and reconciled correctly. A mature go-to-market operation could justify evaluating both.
Pros and Cons for RevOps Teams
ChatGPT Work – Pros:
- Generalist capability means faster deployment across multiple sales and marketing workflows
- Low barrier to adoption for teams already using ChatGPT in day-to-day work
- Cross-application reach reduces context switching for reps
- Strong potential to lift win rate by ensuring reps have better pre-call and post-call context
ChatGPT Work – Cons:
- Generalist tools require careful governance – without clear prompting standards, output quality varies
- CRM integration depth will depend on connector availability, which is still maturing
- Data residency and enterprise security controls are still being scrutinised by procurement teams at large organisations
AWS ERP Agent – Pros:
- Deny-by-default permissions and a separate agent identity make it easier to pass security and compliance reviews
- Measurable ROI is easier to calculate – reduced DSO and manual AR hours are quantifiable
- Purpose-built focus means less configuration and a tighter feedback loop for finance teams
AWS ERP Agent – Cons:
- Narrower scope means it does not address broader sales productivity or pipeline management needs
- Deployment typically requires deeper IT and ERP system involvement, extending implementation timelines
- Best suited to mid-market and enterprise organisations with significant invoice volumes – smaller teams may not see proportional returns
What This Means for Your Stack in 2026
The broader context here is important. Semrush research published this month found that AI is now actively shaping B2B vendor discovery, evaluation, and shortlisting – which means the Ideal Customer Profile conversation has expanded to include how AI systems perceive and surface your company, not just how your reps communicate your value. Buyers are running their own AI-assisted research before they ever speak to a rep. Tools like ChatGPT Work, which sit inside the selling workflow, need to be evaluated alongside that reality.
For RevOps leaders building or refining their stack, the practical recommendation is to separate the evaluation into two tracks. First, assess where AI can reduce friction in the customer-facing revenue motion – prospecting, qualification, follow-up, and pipeline hygiene. ChatGPT Work belongs in that conversation. Second, assess where AI can improve the financial accuracy and speed of revenue recognition – collections, reconciliation, and reporting. AWS’s ERP agent belongs in that conversation. Both tracks ultimately affect Net Revenue Retention (NRR), which is the metric that ties sales efficiency to financial outcomes most directly.
If you are mapping out where these tools might fit alongside your existing CRM infrastructure, the Tool Reviews section at CRM Daily covers an expanding range of AI-enabled sales and revenue tools. You can also browse the CRM Tools Directory to compare options across categories. The agent era is not approaching – it is already here, and revenue teams that build evaluation frameworks now will be better positioned to move quickly as these platforms mature through the second half of 2026.