An autonomous AL extension that monitors a mailbox, identifies the customer and requested items, checks availability, and drafts quotes and sales orders inside Business Central. Built, installed as a per-tenant extension, and supported by ECOSIRE. Built to order by ECOSIRE for Dynamics 365 BC (build-to-order) — indicative price from $799.00 USD; request a quote for a scoped proposal.
应用清单
围绕您的工作流程构建
An autonomous AL extension that monitors a mailbox, identifies the customer and requested items, checks availability, and drafts quotes and sales orders inside Business Central. Built, installed as a per-tenant extension, and supported by ECOSIRE.
无需自行搭建——由 ECOSIRE 构建、安装并提供支持的可用应用。
以一次性构建价格开始。我们在启动时与您共同确定范围。
ECOSIRE 在您的 Dynamics 365 Business Central 上构建、配置并安装。
约 2–4 周内上线,并提供上线后的支持期。
The AI Sales Order Agent turns your inbound order mailbox into structured sales documents inside Microsoft Dynamics 365 Business Central — without rekeying. It is a per-tenant AL extension that ECOSIRE builds, installs and supports on your Business Central environment. This is not an instant AppSource download: we scope it to your item catalog, customer matching rules and approval policy, then deploy it as a signed per-tenant extension you own.
How it works. A scheduled Job Queue entry polls a designated shared mailbox (via Microsoft Graph) for new order emails. Each message is passed to an LLM extraction step that identifies the sender's Customer record (email-to-contact match, with a configurable fallback to manual review), parses requested items against your Item and Item Reference tables, and resolves quantities and units of measure. The agent checks availability using Business Central's inventory and availability data, then creates a Sales Quote and renders the standard quote PDF via your report layout. A new AI Agent Log table (table + page extension) records every decision, the source email, confidence scores and the resulting document number for full auditability.
You stay in control. Low-confidence matches route to a review queue instead of posting blind. On approval, the quote converts to a Sales Order through standard BC document flow — so pricing, discounts, dimensions and posting rules all behave exactly as your finance team configured them. The agent can also manage structured follow-up email exchanges (clarifications, confirmations) tied back to the originating thread.
Everything runs inside your tenant under a scoped permission set. No data leaves your environment except the extraction call to the LLM endpoint you authorize. We deliver the source-controlled AL app, install it, and support it.
Owns inbound order-taking and wants emailed orders turned into BC quotes and sales orders without staff rekeying, while keeping a human approval gate on anything uncertain.
Responsible for the tenant and needs a clean per-tenant extension with a scoped permission set, Job Queue setup and an audit log — not an opaque black box bolted onto production.
Manages the order desk inbox and wants the agent to handle the repetitive parsing and follow-up clarification emails so the team focuses on exceptions and edge cases.
| 标准 | 伊科西尔 | 定制建造 | 竞争对手 | 奥杜本机 |
|---|---|---|---|---|
| Reads inbound order emails and a designated mailbox automatically | 包含 | 部分支持 | 部分支持 | |
| AI extraction of customers and items from free-text and PDF attachments | 包含 | 部分支持 | 部分支持 | |
| Drafts quotes and converts to sales orders via native BC document flow | 包含 | 包含 | 部分支持 | |
| Tailored to your item catalog, customer-matching and approval rules | 包含 | 包含 | 不包括在内 | |
| Human review queue with configurable confidence thresholds | 包含 | 部分支持 | 部分支持 | |
| Full audit log of every decision and source email | 包含 | 部分支持 | 部分支持 | |
| Installed as a per-tenant extension you own, with source code | 包含 | 包含 | 不包括在内 | |
| Built, installed and supported for you (no in-house AL team needed) | 包含 | 不包括在内 | 部分支持 |
This is a build-to-order extension, not an instant AppSource download. After a short scoping call we typically deliver in 3-5 weeks depending on the complexity of your item catalog and customer-matching rules. We build the AL app, install it first on your sandbox as a signed per-tenant extension, validate against your real emails, then deploy to production.
Every build includes an agreed warranty window for bug fixes. Because it is a per-tenant extension you own (with the AL source in your repository), we also offer ongoing support for Business Central version upgrades, new item/customer rules and extraction-prompt tuning. We track BC release-wave compatibility so the extension keeps installing cleanly through major updates.
No data leaves your tenant except the text extraction call to the LLM endpoint you authorize. Mailbox access uses your own Microsoft Graph app registration, and all document creation happens inside Business Central under a scoped permission set. Every decision is recorded in the AI Agent Log table for audit.
Only if you configure it that way. By default, low-confidence customer or item matches are held in a review queue, and the human-approved quote-to-order conversion uses standard BC document flow so your pricing, discount, dimension and posting rules all apply. You set the confidence thresholds.
Yes. The extraction step handles both free-text email bodies and PDF/spreadsheet attachments, and resolves items against your Item Reference and customer-specific cross-reference numbers. Unmatched lines are flagged for review rather than guessed.
It is built primarily for Business Central online (SaaS) as a per-tenant extension using Job Queue, the Graph API and the BC platform APIs. On-premises deployments are possible — we confirm feasibility and any platform constraints during scoping.
An autonomous AL extension that monitors a mailbox, identifies the customer and requested items, checks availability, and drafts quotes and sales orders inside Business Central. Built, installed as a per-tenant extension, and supported by ECOSIRE.