--- url: 'https://chatfin.ai/blog/what-we-learned-connecting-ai-agents-to-sap-business-one/' title: What We Learned Connecting AI Agents to SAP Business One author: name: Tara Parker url: 'https://chatfin.ai/author/chatfin_admin/' date: '2026-06-04T16:35:20+00:00' modified: '2026-09-28T15:57:05+00:00' type: page summary: 'What actually broke when we connected AI agents to SAP Business One: user defined fields, add on tables, Service Layer sessions and the reporting workarounds.' published: true --- # What We Learned Connecting AI Agents to SAP Business One - Blog · SAP Business One & AI # What We Learned Connecting AI Agents to SAP Business One By [Ashok Manthena](https://www.linkedin.com/in/ashokman/) B1's finance depth is real. The problems were all in the places each customer had customised. AM [ Ashok Manthena ](https://www.linkedin.com/in/ashokman/) - Finance AI Architect, ChatFin Published April 27, · 9 min read · SAP B1 & AI In This Article [ SAP B1's connection architecture: DI API vs Service Layer ](#connection-architecture) - [ Setting up the SAP B1 Service Layer for AI ](#service-layer-setup) - [ Lesson 1: The Service Layer has entity limits per query ](#lesson-1) - [ Lesson 2: Multi-branch SAP B1 is significantly more complex ](#lesson-2) - [ Lesson 3: Month-end close is the highest-ROI use case ](#lesson-3) - [ Lesson 4: AP automation in SAP B1 - what the numbers show ](#lesson-4) - [ Lesson 5: Data normalization matters more than the API ](#lesson-5) - [ Lesson 6: The finance team is the last integration point ](#lesson-6) - [ FAQ ](#faq) TL;DR - Key Takeaways - SAP B1 Service Layer is the right connection path for AI - not the older DI API - The close cycle dropped from 72.5 hours to under 3 hours for one SAP B1 customer shown in the webinar - AP automation: 147 invoices processed overnight with AI agents connected to SAP B1 - SAP B1's chart of accounts structure is often inconsistent across branches - clean this before connecting AI - Most teams go live in 2 weeks without engaging an external SAP consultant SAP Business One sits in a distinctive position in the ERP landscape. It is not the enterprise behemoth that SAP S/4HANA is, yet it carries far more finance depth than most lightweight accounting platforms. For the 80,000-plus companies running their operations on SAP B1, the question is no longer whether AI belongs in the finance stack - it is how to connect it cleanly, maintain data integrity, and build the kind of trust that gets controllers to sign off on autonomous overnight runs. ChatFin has connected AI agents to SAP Business One across dozens of finance teams, from single-entity manufacturers to multi-branch distributors with operations across three continents. What we encountered was not a technology problem - the SAP B1 Service Layer is genuinely well-built. The challenges were architectural, organizational, and deeply tied to years of accumulated data decisions that no one had ever had to surface before. This article is the honest account of what we learned. ## SAP B1's Connection Architecture: DI API vs Service Layer SAP Business One has two primary APIs, and choosing the wrong one will cost you weeks. The older path is the ** DI API (Data Interface API) ** - a COM-based interface that has been part of SAP B1 since its earliest versions. The DI API requires a Windows environment, a dedicated SAP B1 client installation on the connecting machine, and a series of COM object registrations that make it essentially incompatible with cloud-native AI agent architectures. It was designed for tightly coupled integrations in a Windows-first world, and it shows. The modern path is the ** Service Layer ** , introduced with SAP Business One 9.x and substantially matured through version 10.0. The Service Layer exposes SAP B1 data and business objects through a REST/OData interface that runs over HTTPS. It is platform-agnostic, can be consumed from any language or runtime that can make HTTP requests, and follows standard OData query conventions that AI agents can reason about natively. ChatFin uses the Service Layer exclusively. | Dimension | DI API | Service Layer | | --- | --- | --- | | Transport | COM / Windows RPC | REST / OData over HTTPS | | Platform requirement | Windows only | Any platform | | SAP B1 client required | Yes | No | | Suitable for cloud AI agents | No | Yes | | Authentication | SAP B1 client credentials | Session token or NTLM | | Base endpoint pattern | COM object registration | https://[server]:50000/b1s/v1/ | The Service Layer endpoint pattern follows a consistent structure: ` https://[your-server]:50000/b1s/v1/[EntitySet] ` . Authentication is handled via a session-based login that returns a session token, or alternatively through NTLM. Once authenticated, your AI agent can query business objects - JournalEntries, BusinessPartners, PurchaseInvoices, Accounts, and dozens more - using standard OData filters, selects, and ordering. The response comes back as JSON, which means AI agents can parse and reason about it without any additional transformation layer. ## Customisation is where it breaks B1 finance depth holds up well. The problems appear in the places each customer configured differently. > I'm > a > big > proponent > of > trial > and > error. > I > use > it > at > home, > I > use > it > at > work. > The > best > way > to > learn > how > to > use > it > for > work > is > to > use > it > for > a > lot > of > different > things. > Adam > Swoverland, > on > the > AgentCFO > podcast Inventory your own B1 customisations before integration work starts, and test the agent against those first. ## Setting Up the SAP B1 Service Layer for AI The setup process is straightforward once you know the sequence. Here is what we walk every SAP B1 customer through in the first session: 1 ** Enable the Service Layer in SAP B1 Admin. ** Navigate to the SAP Business One Service Layer configuration utility on your server. If Service Layer is not running, the setup wizard will walk you through the initial configuration. Verify the service is active on port 50000 (default) and confirm HTTPS is enabled. 2 ** Configure SSL certificates. ** SAP B1 Service Layer ships with a self-signed certificate. For production AI agent connections, replace this with a certificate from a trusted CA or configure your agent to trust the self-signed certificate explicitly. Skipping this step causes intermittent SSL handshake failures that are difficult to diagnose. 3 ** Create a dedicated API user account. ** This is the step most teams get wrong. Do not use an existing finance user's credentials, and absolutely do not use an admin account. Create a new SAP B1 user specifically for API access with read permissions on the relevant business objects (GL accounts, business partners, invoices, journal entries). Write permissions are added later during the write-back UAT phase. 4 ** Test the login endpoint. ** The login call goes to ` POST /b1s/v1/Login ` with a JSON body containing CompanyDB, UserName, and Password. A successful response returns a session ID that the AI agent includes in all subsequent requests as a cookie or header. 5 ** Validate key entity endpoints. ** Pull a sample from ChartOfAccounts, BusinessPartners, and JournalEntries to confirm data is accessible and the field names match what your finance team uses for reporting. This validation step catches permission gaps before the AI agent hits them at runtime. The most common mistake we see is using an admin SAP B1 account for API connections. Beyond the obvious security risk of exposing admin credentials in a configuration file, admin accounts often have access to system-level operations that the AI agent should never touch. A least-privilege API user is not optional - it is the foundation of a safe integration. ## Lesson 1: The Service Layer Has Entity Limits per Query The first thing that surprises finance teams when they start pulling data at scale is that the SAP B1 Service Layer defaults to returning ** 20 entities per query ** . This is a deliberate design choice to protect server resources, but it creates a problem the moment you try to pull AP aging data across 500 vendors or retrieve a full GL transaction history for a fiscal quarter. The solution is OData-standard pagination using the ` $top ` and ` $skip ` parameters. A query for all open purchase invoices might look like this: ` GET /b1s/v1/PurchaseInvoices?$filter=DocumentStatus eq 'bost_Open'&$top=100&$skip=0 GET /b1s/v1/PurchaseInvoices?$filter=DocumentStatus eq 'bost_Open'&$top=100&$skip=100 GET /b1s/v1/PurchaseInvoices?$filter=DocumentStatus eq 'bost_Open'&$top=100&$skip=200 ` ChatFin's AI agent handles this pagination automatically. The agent detects when a result set is paginated, assembles the full dataset before reasoning about it, and ensures that finance calculations - aging buckets, accrual totals, reconciliation variances - are based on complete data rather than the first page. Teams that build their own integrations without handling pagination end up with AI agents that are confidently wrong: they produce numbers that look plausible but are based on a 20-record sample of a 600-record dataset. Beyond pagination, the ` $select ` parameter is important for performance. The Service Layer will return all fields on an entity by default, many of which are irrelevant to finance queries. Selecting only the fields you need reduces payload size significantly and speeds up the agent's reasoning cycle. For AP aging, that typically means DocNum, CardCode, CardName, DocDate, DocDueDate, DocTotal, PaidToDate, and DocumentStatus - not the full 80-field invoice object. ## Lesson 2: Multi-Branch SAP B1 Is Significantly More Complex SAP Business One's multi-branch architecture is one of the more consequential things to understand before you start an AI integration. Unlike some ERPs that handle multiple entities within a single database using branch or segment flags, ** SAP B1 typically separates each company entity into its own company database ** . A holding company with three operating subsidiaries will have three separate SAP B1 company databases, each with its own chart of accounts, business partner master, and transaction history. This is not a flaw - it is an architectural choice that gives each entity genuine data separation. But it means AI agents must be configured with separate connections for each company database, and consolidated reporting requires the agent to query multiple databases, normalize the results (accounting for the fact that GL account codes may differ between entities), and assemble a coherent consolidated view. Each company database requires its own authentication session and Service Layer endpoint configuration Chart of accounts codes are rarely identical across entities - the agent needs a mapping layer or relies on account names Intercompany eliminations must be handled explicitly - the AI agent cannot infer them from the data structure Consolidated P&L and balance sheet generation requires currency conversion for entities in different functional currencies Reporting latency increases proportionally with the number of entities - typically still under 60 seconds for three-entity consolidations This is consistently the most common implementation surprise for multi-branch SAP B1 customers. The single-entity setup takes a day. The multi-entity setup takes a week, most of which is spent on the account mapping exercise rather than the technical connection itself. Teams that account for this in their project timeline stay on schedule. Teams that assume multi-branch is a simple extension of single-entity - do not. ## Lesson 3: Month-End Close Is the Highest-ROI Use Case Of all the finance processes that AI agents can automate in SAP Business One, ** month-end close delivers the most dramatic and most immediately measurable improvement ** . The webinar above shows a real customer who reduced their close cycle from 72.5 hours to under 3 hours. That is not a projected figure - it is a measured outcome from a production deployment running overnight close procedures autonomously. The SAP B1 month-end close sequence involves a series of discrete, sequential procedures that are well-documented and deterministic: the period-end closing wizard, AR reconciliation, AP reconciliation, currency revaluation for foreign-currency accounts, depreciation runs, inventory revaluation if relevant, and finally a trial balance generation. Each step has clear inputs, clear success criteria, and clear outputs. This is exactly the kind of workflow that AI agents execute reliably. ChatFin's month-end close agent for SAP B1 works through the sequence overnight. It begins once the business day closes, runs the period-end wizard through the Service Layer, and queries AR and AP balances to flag unreconciled items above a materiality threshold. It then triggers currency revaluation using the rates configured in the SAP B1 Exchange Rate table, runs the depreciation calculation check, pulls the trial balance, and generates a close package summary - including any exceptions that require human review - that lands in the controller's inbox before 7am. Learn more about how [ AI for month-end close ](https://chatfin.ai/blog/ai-for-month-end-close-and-reconciliations-close-in-days-not-weeks/) works across ERP platforms. 72.5h Previous close cycle duration (SAP B1 customer) <3h Close cycle with AI agents running overnight 27/30 Days closed without a human in the close sequence "SAP Business One has more finance depth than most people realize. The challenge isn't the ERP - it's making the AI match the finance team's mental model of the data." - Ashok Manthena, ChatFin The month-end close use case also surfaces something important about where AI agents add value in SAP B1 specifically. The ERP already enforces most of the control logic - period locking, posting period validation, document numbering. The AI agent does not need to replicate those controls. It operates within them, using the Service Layer to trigger the same procedures a controller would run manually, but doing so systematically, at scale, and at a time when human resources are not available. For more on this pattern, see our guide to agentic finance for CFOs and controllers . ![ChatFin month-end close automation: from 72.5 hours to under 3 hours with AI agents connected to SAP Business One](/wp-content/uploads/2026/05/chatfin-month-end-close-receipt-72-hours.png) * From 72.5h to under 3h: SAP B1 month-end close with ChatFin AI agents running overnight procedures autonomously. Close package delivered to controller inbox before 7am. * ## Lesson 4: AP Automation in SAP B1 - What the Numbers Show The second highest-ROI use case after month-end close is accounts payable automation. The numbers from one SAP B1 deployment: ** 147 invoices coded and posted overnight ** without human intervention. That was a Tuesday night run during a peak vendor payment cycle. The AP team arrived Wednesday morning to a queue that had already been processed, coded, and staged for payment approval. The SAP B1 AP workflow follows a specific sequence that maps cleanly onto AI agent task orchestration. An incoming purchase invoice in SAP B1 requires vendor identification (mapping the incoming invoice header to a Business Partner record), document creation (posting the purchase invoice with the correct document series), GL coding (assigning the expense or asset account based on the line item description and the vendor's default account), and optionally a payment run entry. The AI agent handles all three document steps through the Service Layer. ** Vendor matching: ** The agent queries the BusinessPartners entity using fuzzy name matching and tax ID cross-referencing to identify the correct BP record, even when the invoice header uses abbreviated or alternate vendor names ** GL coding: ** Line item descriptions are matched against the chart of accounts using a learned coding pattern trained on the previous 12 months of manual entries - the AI adapts to the company's specific coding conventions ** Exception routing: ** Invoices with ambiguous GL assignments, duplicate detection flags, or amounts above the auto-approval threshold are routed to the AP manager with a pre-populated coding suggestion rather than a blank form ** 80% reduction in manual AP touches: ** The AP team shifts from data entry to exception review - a fundamentally different and higher-value use of their time For teams evaluating this capability, the full analysis of what AI-driven AP automation delivers in practice - including cost-per-invoice benchmarks and implementation sequencing - is covered in our [ complete guide to building AI agents for accounts payable automation ](https://chatfin.ai/blog/building-ai-agents-for-accounts-payable-automation-complete-guide/) . The AP/AR use case page has the summary view with ROI calculators. For cost benchmarking specifically, see how AI reduces invoice processing costs by 80% . ## Lesson 5: Data Normalization Matters More Than the API Here is the lesson that comes as the biggest surprise to teams with strong technical resources: the API connection is not the hard part. The API connection takes hours. ** Data normalization takes days to weeks ** , and it determines 80% of the quality of what your AI agents can produce. SAP Business One installations accumulate data quality debt over time, particularly in the areas that matter most for AI reasoning: the chart of accounts, the business partner master, and dimension tracking. A company that has been on SAP B1 for six years typically has some combination of the following issues: ** Duplicate business partners: ** The same vendor exists under three different names ("Acme Corp", "ACME Corporation", "Acme Corp.") with transactions split across all three. AP aging and vendor spend analysis are unreliable until these are merged. ** Uncategorized GL entries: ** A catch-all "miscellaneous expense" account that contains anything from legitimate miscellaneous items to entries that should have gone to specific cost categories. AI agents cannot reclassify these without a human decision on each. ** Missing cost centers / dimensions: ** SAP B1 has a flexible dimension tracking system, but teams often start using it inconsistently mid-implementation. A year of transactions have no dimension codes; the next year has partial codes; the third year is clean. Financial reporting by department or project becomes incoherent across that window. ** Inconsistent account codes across branches: ** In multi-entity setups, account code 4010 is "Product Revenue" in one entity and "Service Revenue" in another. Consolidated P&L requires a manual mapping table that no one has ever formalized. ** Inactive records cluttering queries: ** Dozens of inactive vendors, closed bank accounts, and deprecated GL accounts that still appear in queries because they were never properly archived in SAP B1. The data audit we run before connecting AI to SAP B1 systematically surfaces all of these issues. We query business partner counts and flag duplicates by tax ID. We pull GL entry distributions by account and flag accounts with high variance in entry descriptions. We check dimension completeness rates by fiscal period. This audit report goes to the finance team before the AI agent goes live - because the finance team has to own the cleanup decisions. The AI can surface the problems, but it cannot make the judgment call on how to resolve a six-year-old vendor duplicate that has payments attached to both records. Teams that invest two to four days in data normalization before go-live produce dramatically better AI outputs from day one. Teams that skip this step spend weeks debugging AI responses that are technically accurate but financially misleading because the underlying data is inconsistent. See also our [ step-by-step guide to building AI agents for reconciliations ](https://chatfin.ai/blog/step-by-step-guide-building-ai-agents-for-reconciliations/) for how data quality affects reconciliation accuracy specifically. ## Lesson 6: The Finance Team Is the Last Integration Point The technology is typically ready in two to three days. The Service Layer connection is established, the agent is querying SAP B1 correctly, the data normalization work is scoped and underway. And then the project slows down - not because of a technical blocker, but because the finance team, reasonably, wants to understand what they are signing off on before they trust an AI agent with their books. This is not resistance to change. It is professional judgment. Controllers and finance managers are personally responsible for the accuracy of the financial records in SAP B1. They have audit exposure. They have regulatory obligations. The idea that an AI agent is going to post journal entries, reconcile accounts, and generate board-level reporting without their direct involvement requires a level of demonstrated reliability that no vendor can assert - it has to be earned. The deployment sequence that works consistently is incremental trust-building: 1 ** Read-only dashboards first (Week 1 to 2). ** The AI agent queries SAP B1 and surfaces GL summaries, AP aging, and open item reports in a dashboard format. Finance sees what the AI sees. They validate that the numbers match what they pull manually. Trust begins. 2 ** Accrual calculations with human approval (Week 3 to 4). ** The AI agent proposes accrual entries based on the open purchase orders and unbilled items in SAP B1. The controller reviews, approves, and posts manually. The agent's calculations are audited against the controller's judgment. 3 ** AP coding with exception workflow (Week 4 to 6). ** The AI agent codes incoming invoices and routes exceptions. The AP team reviews the coded invoices before posting. Accuracy is measured. The approval threshold is adjusted based on observed error rates. 4 ** Overnight close automation (Week 6 to 8). ** Once the finance team has seen the agent operate correctly across 30-plus close procedures in supervised mode, the overnight autonomous close run is enabled. The controller sets the exception thresholds. The agent handles everything within those bounds and flags everything outside. The finance team is not the last obstacle to AI integration - they are the last integration point. Their domain knowledge about how the chart of accounts is actually used, which vendors are problematic, which GL codes have known ambiguities, and which month-end adjustments are judgment calls rather than mechanical calculations is what makes the AI agent genuinely useful rather than theoretically capable. For a deeper look at this dynamic, see our piece on what agentic finance means for CFOs and controllers in . ## FAQ: AI Agents and SAP Business One What API does ChatFin use for SAP Business One? ChatFin uses the SAP B1 Service Layer - the modern REST/OData API available in SAP B1 version 9.x and later. The older DI API is COM-based and Windows-only, making it unsuitable for cloud-native AI agent architectures. All queries, data retrieval, and write-back operations go through the Service Layer endpoint at ` https://[server]:50000/b1s/v1/ ` . Does ChatFin work with SAP Business One 10.0? Yes. SAP B1 10.0 has the most current Service Layer implementation, making it the easiest version to connect to AI agents. All Service Layer endpoints are stable and well-documented in version 10.0, and the authentication and entity model are fully compatible with ChatFin's agent architecture. Teams on 9.2 and 9.3 also work, with minor adjustments for entity availability differences. How is multi-branch SAP B1 handled? Each company database connects separately. ChatFin's agents can run cross-database queries for consolidated reporting, but each connection must be configured individually with its own authentication credentials and Service Layer endpoint. The data normalization work - particularly chart of accounts mapping across entities - is the primary time investment in multi-branch implementations. This is consistently the most complex aspect of SAP B1 AI integration. What does the SAP B1 month-end close automation look like? The AI agent runs the period-end closing wizard, reconciles AR and AP balances, flags unreconciled items above a configurable materiality threshold, handles currency revaluation using rates from the SAP B1 Exchange Rate table, runs the depreciation calculation check, generates a trial balance, and assembles a close package with exception flags - all overnight, without human intervention. The controller receives the close package summary with a clear exception list by 7am. Is write-back available for SAP B1? Yes, with appropriate permissions and finance team UAT. Write-back through the Service Layer supports journal entry posting, payment application, invoice status updates, and business partner record maintenance. Write-back is enabled incrementally as the finance team validates AI agent accuracy in read-only mode first. Admin permissions are never required - a least-privilege API user with specific write scopes is always the correct configuration. Related Reading - Use Case [ AI for Month-End Close - ChatFin Use Case Overview ](https://chatfin.ai/blog/ai-for-month-end-close-and-reconciliations-close-in-days-not-weeks/) - Use Case AI for AP/AR - ChatFin Use Case Overview - Guide [ Step-by-Step Guide: Building AI Agents for Reconciliations ](https://chatfin.ai/blog/step-by-step-guide-building-ai-agents-for-reconciliations/) - ERP [ How to Connect JD Edwards to Claude AI ](https://chatfin.ai/blog/how-to-connect-jd-edwards-to-claude-ai/) - ERP [ AI Agents for Acumatica Finance Teams ](https://chatfin.ai/ai-agents-for-acumatica-finance-teams/) - AP [ Complete Guide: Building AI Agents for AP Automation ](https://chatfin.ai/blog/building-ai-agents-for-accounts-payable-automation-complete-guide/) - External [ My Journey Connecting AI Agents to SAP Business One (Medium) ](https://medium.com/@tara_51063/my-journey-connecting-ai-agents-to-sap-business-one-7d48a92e30b9) ## The Bottom Line: SAP Business One Is Ready for AI SAP Business One is genuinely well-positioned for AI integration. The Service Layer API is stable, well-documented, and capable of handling the query patterns that AI agents need for finance automation. The data model is consistent and the finance processes - month-end close, AP, reconciliations - are standardized enough that AI agents can execute them reliably once configured. The lesson from every SAP B1 AI implementation is this: ** start with data quality, set honest expectations with the finance team, and let the AI earn trust incrementally. ** Do not connect the API and immediately try to automate everything. Run the data audit first. Fix the business partner duplicates. Clarify the miscellaneous GL accounts. Build the month-to-month account mapping for multi-entity consolidations. Then connect the AI and start with read-only reporting before any autonomous posting goes live. Teams that follow this sequence go live in two weeks and see measurable ROI within the first close cycle. Teams that skip the data quality step spend that same two weeks debugging outputs that are technically correct but financially confusing. The API is the easy part. The data and the trust are the work. [ Talk to us about connecting AI agents to your SAP Business One instance → ](https://chatfin.ai/talk-to-us/) Free Demo See ChatFin on your SAP B1 data 15-minute live demo. We connect to your SAP B1 Service Layer and run a month-end close query on the spot. [ Book a Demo → ](https://chatfin.ai/talk-to-us/) ![ChatFin month-end close efficiency metrics: 72.5 hours to under 3 hours with SAP Business One AI agents](/wp-content/uploads/2026/05/chatfin-month-end-close-receipt-72-hours.png) SAP B1 AI Results ### By the Numbers 72.5h → <3h Month-end close cycle reduction 147 Invoices coded overnight (AP automation) 80% Reduction in manual AP touches 2 weeks Typical go-live timeline (no SAP consultant needed) ERP Integrations ### ChatFin Supports - Current Page [ SAP Business One ](https://chatfin.ai/ai-erp/what-we-learned-connecting-ai-agents-to-sap-business-one/) - Guide [ JD Edwards (E1 & World) ](https://chatfin.ai/blog/how-to-connect-jd-edwards-to-claude-ai/) - Guide [ Acumatica ](https://chatfin.ai/ai-agents-for-acumatica-finance-teams/) - Podcast [ ChatFin Podcast - Finance AI Series ](https://chatfin.ai/podcast/) Use Cases ### Finance Automation - Close [ AI for Month-End Close ](https://chatfin.ai/blog/ai-for-month-end-close-and-reconciliations-close-in-days-not-weeks/) - AP/AR AI for AP & AR - FP&A [ AI for FP&A ](https://chatfin.ai/blog/ai-for-fpa-budgeting-and-forecasting-automation/) - Accruals [ AI for Accruals ](https://chatfin.ai/blog/ai-for-accruals-automate-accrual-calculations-and-journal-entries/) Further Reading ### Related Posts - [ Top 10 AI Tools for CFOs & Finance Leaders ](https://chatfin.ai/blog/top-10-best-ai-tools-for-cfos-finance-leaders/) - [ AP Automation: CFO Decision Guide for ](https://chatfin.ai/blog/accounts-payable-ai-automation-cfo-decision-guide-for-2026/) - [ The Missing Pillar in FP&A - ChatFin Podcast ](https://chatfin.ai/blog/chatfin-podcast-the-missing-pillar-in-fpa-with-the-fpa-guy/) ## Where SAP Business One reporting runs out SAP Business One handles transactions well. These five gaps are documented, and every one of them ends in a spreadsheet. **No native consolidation.** Multi company consolidation means exporting each company's reports and combining them in Excel, which is why close week reruns the same pack repeatedly. **No transaction currency reporting.** Native reports cannot produce statements in transaction currencies, which makes reconciliation harder and slows period end. **One accounting standard per company.** B1 assumes a single reporting standard per managed company, so running US GAAP and IFRS in parallel means manual adjustment at consolidation. **Rigid report formats.** A chart of accounts change forces manual restructuring of income statement rows rather than the report adapting. **Weak statutory mapping.** You can hold an external code per GL account, but there is no mapping to multiple alternate charts and no way to run statements on them. An agent connected over the Service Layer reads across companies and answers the consolidated question directly, which is the step that currently happens in Excel.