Summarise the article with your AI
Here you can read:
Treasury tools have always required someone to leave their workflow, open a platform, and come back with an answer, by which point a payment is still waiting, or a cash balance has already moved. The Embat-Slack integration removes that step entirely: it doesn't just surface treasury data in Slack, it lets the team act on it.
The CFO gets a Slack message with a payment batch ready for approval: total amount, number of payments, who prepared it. She approves it from her phone between meetings, without logging into a platform she rarely opens. That's the shift this integration is about.
The Embat-Slack integration, powered by TellMe operating in Ask mode, brings treasury management, queries and actions alike, directly into Slack and helps your team collaborate effectively.
Gartner's 2025 survey found that knowledge management (the art of organising, retrieving, and using information for better decisions) is now the most common AI use case in finance organisations (49%). Conversational AI makes complex financial systems more accessible for non-specialists whilst maintaining the security and governance controls treasury actually requires.
How Embat connects to Slack and how it works
Embat makes exactly this possible: It brings the full treasury management platform into Slack via TellMe's Ask mode, allowing finance teams to query live financial data and, critically, execute treasury operations without leaving their workspace.
The Embat-Slack connector is not simply an AI chatbot, it is the Embat treasury platform itself, via TellMe, Embat's AI agent, surfaced inside the communication tool the team already uses. Same data, same permissions model, same action layer. When someone asks a question in Slack, embat pulls live data directly from the treasury infrastructure and returns a cited answer, and when someone needs to act, Embat executes it, with a human confirming every step.
Three ways to talk to Embat in Slack
The integration offers three interaction modes tailored to your team's privacy and visibility requirements:
1. Channel mentions: when the whole team needs the answer
A @Embat mention in a shared #treasury channel works when the information, or the decision, is collectively useful, like with end-of-day cash positions, a confirmation that a payment batch cleared or a flagged anomaly in an overnight transaction. It's also how a CFO approves a payment batch in front of the team: the batch appears with full context, and one click in the channel approves or rejects it, visible to everyone. Everyone gets the same answer from the same source at the same moment, thereby eliminating the "someone mentioned X in a different thread" problem that can cause misjudgements.
Such channel-visible answers directly address primary adoption barriers: Research from the Association of Corporate Treasurers shows that cultural barriers remain the primary obstacle to AI adoption in UK treasury functions, with 50% of treasurers admitting reluctance to change existing manual methods despite 95% of CFOs feeling at least somewhat prepared to adopt AI. Making AI output transparent and collectively verifiable helps normalise its use.
2. Ephemeral slash commands: your question, only your answer
The /Embat slash command returns an answer visible only to the person who asked. A head of treasury checking, for example, whether a specific entity is below its minimum cash threshold before joining a board call would use this mode. It's also the mode for a one-off action nobody else needs to see: "convert 200K EUR to USD in my Embat wallet" triggers the FX conversion privately, at the moment the rate looks favourable. The answer, or the action, appears, the decision gets made, and no message trail is left in the channel, with permission-aware confidentiality built into the workflow.
3. Direct messages: for extended analysis without the channel noise
Direct Message (DM) mode with Embat also suits iterative queries like starting with a cash position summary, drilling into a specific entity, and then asking for a breakdown by currency. It's also where a treasurer can walk through a cash optimisation proposal step by step, reviewing the suggested movements between entities and accounts, asking follow-up questions, and confirming the execution once satisfied. Because Slack DMs persist like a conversation thread, the analysis is preserved and reviewable, which proves useful for preparing board reports or month-end reviews.
From answering questions to taking action
This is the core of the Embat-Slack integration. TellMe was built from the start to handle both retrieval and action, not just answering questions but enabling a team to act. In Slack, that means a question about pending payments can become a reviewed, confirmed action without opening a separate tool.
The integration operates exclusively in Ask mode, meaning every query and every action is explicitly initiated by a human. TellMe is designed to not act autonomously, does not push unsolicited alerts, and does not execute anything without confirmation. Treasury operations require human judgement, and the integration is designed to support that judgement, not replace it.
Four actions show what this looks like in practice:
1. Approving payments. Payment batches are usually prepared by the treasury team, yet approval sits with the CFO or Finance Director profiles that don't work in Embat day to day. For them, approving today means logging into a tool they rarely open just to click a button. In Slack they get the batch with full context: total amount, number of payments, who prepared it, and approve or reject in one click, from desktop or mobile. Approval stops being the bottleneck of the payment cycle.
2. Creating Embat wallets and executing FX between wallets. Embat gives customers multi-currency wallets to hold and pay in local currencies. From Slack, a treasurer could open a new currency wallet the moment the need appears (say, first supplier invoicing in PLN) or execute an FX conversion between their own wallets when the rate is favourable. These are one-off, decision-driven operations that fit a conversational interface far better than a full platform session.
3. Running cash optimisation. Idle balances spread across entities and accounts translate directly into bank costs, paying overdraft interest on one account while cash sits unused in another. From Slack, Embat can propose the optimal set of movements and execute it after a single confirmation. This is the action with the most tangible ROI: measurable savings in bank costs triggered from a message.
4. Uploading files. Finance teams already receive files through Slack: bank statements from banks without direct connectivity, amortisation schedules from lenders, remittance advice, and many more. Today those must be downloaded and re-uploaded into Embat. With the integration, you drop the file into the Embat conversation and it gets processed in Embat. This way, Slack becomes an ingestion channel instead of a place where files pile up.
Comparing Embat modes in Slack:
| Mode | What it does | Who sees it | Best used for |
|---|---|---|---|
| Channel @mention | Pulls live treasury data into channel and enables shared approvals | Everyone in channel | Team alignment and visible sign-off |
| Slash command | Returns data or executes an action only to user | Only querying user | Pre-meeting checks, sensitive queries and one-off actions |
| Direct message | Full conversational thread for analysis and step-by-step confirmations | Only querying user | Iterative analysis, report prep and reviewing complex actions |
For a detailed comparison of how Embat's conversational treasury approach works across different AI interfaces, see our Claude integration article.
Security and data access: what finance teams need to know
Any CFO or IT security lead will typically evaluate a finance integration on governance first, because treasury data security is not negotiable. Three specific concerns tend to shape that evaluation:
Permissions are enforced at the data level, not the interface level
A user querying Embat through Slack can only access the companies, entities, and modules they're already authorised to see in the treasury platform. The integration doesn't create a new access layer, but inherits the existing one. If someone asks Embat a question about an entity they're not permitted to see, Embat won't return that data. The same principle applies to actions: a user can only approve, convert, or execute what their existing permissions in Embat already allow.
OAuth authentication: no new credentials
The connection between Slack and Embat is established once via OAuth, the same standard used for example for "Signing in with Google". Embat does not store or share the user's Slack credentials, and Slack does not receive treasury data in permanent storage. Embat is only available in channels it's been explicitly invited to and doesn't monitor other conversations, DMs between colleagues, or channels it hasn't been added to.
In addition, Embat's ISO 27001 certification and SOC 2 Type II attestation provide the enterprise-grade assurance.
No data leaves the treasury infrastructure
Embat in Slack surfaces data, it doesn't export it. When Embat answers a query or confirms an action in a Slack channel, the response comes from a live, permissioned API call to the treasury infrastructure. The data is not stored in Slack, not cached by a third-party intermediary, and not used to train any model.
How this changes the way finance teams actually work
When cash positions, payment statuses, and reconciliation flags live in a platform that only some people check regularly, the rest of the team risks making decisions on stale, secondhand information, missing alerts entirely, or waiting on someone else to log in and act.
The Embat–Slack integration solves this by moving treasury data and treasury execution into a shared, queryable, and auditable layer within the communication tool your entire team already uses. Even as PwC's 2025 Global Treasury Survey indicates that 74% of treasurers are expanding or actively using AI, actual value comes from meeting teams where they work and letting them act from there, making treasury management in Slack a natural evolution.
Getting started with Embat in Slack
Setup is a simple, one-time administrative action. A Slack workspace admin completes the connection in Embat's Group Settings. Next, Embat gets invited to relevant channels via /invite @Embat, and the team can start using @Embat, slash commands, or DMs immediately to ask, and to act.
With Microsoft Teams support also on the roadmap, Embat is bringing treasury management into these communication hubs to foster genuine alignment through shared, collective insights.
