Every domain. Every agent. One point of contact.
AI + human customer conversations, answered in Microsoft Teams.
One Teams-native chat platform for AI and human agents. Hosted, or Self-Hosted in your subscription.
- Point of Contact — Hosted:
- The full platform, today.
- Point of Contact — Self-Hosted:
- The shipped core today — one widget (Growth+), AI first, claim and close in Teams, cost dashboard, signed updates; rows marked Coming to Self-Hosted are on their way to the same isolated runtime.
Coming to the Azure Marketplace · Microsoft Teams + Azure required · Two ways to run it
Built on Microsoft Teams and Azure
Microsoft Teams Entry price
From $49/month + your Azure at cost
Point of Contact — Hosted: From $49/month. Same seat ladder on both products.
Workspace
One Teams tab
Self-Hosted today: list, claim, reply and close, with Teams SSO.
The AI
Your Foundry, your tokens
Point of Contact — Hosted: Your agents, your tokens. One Foundry assistant on Self-Hosted today, with an editable system prompt.
Volume
Unlimited AI conversations
No per-resolution fee on any plan.
Fragmented widgets per domain, someone else's tokens, cold handoffs, and anonymous-or-nothing visitors — four ways external chat breaks before a person ever answers.
Three layers, one platform
The platform stacks three layers. Widgets: a branded entry point on every website, brand or app you serve. Flexible AI and human interaction: Foundry agents you own, Teams users, and handoff between them as often as a conversation needs, with hours, presence and a fallback contact. External users and files: registration through your own Entra External ID and secure file spaces for review. All of it in one Teams tab, on Azure AI Foundry you own.
Foundation
Underneath all three pillars: Microsoft Teams — agents and admins in one tab, SSO, activity-feed alerts — and Azure: the Foundry project you own, Blob storage for files, Entra External ID. Secure, scaling with full cost control, easy to configure, and two ways to run it.
- Teams
- Agents and admins in one personal tab, with Teams SSO and activity-feed alerts.
- Azure
- Azure AI Foundry you own, Azure Blob for files, and Microsoft Entra External ID for registered users.
- Secure
- Self-Hosted and Private run in the Azure subscription you deploy them into; Hosted states plainly where the record lives. See Security.
- Scales with full cost control
- A per-seat licence, unlimited AI conversations, and inference billed by Microsoft at Microsoft rates.
- Easy to configure
- Widgets, Foundry agents, groups, hours, sign-in, files and notifications are settings in the Teams tab — changed live, no redeploy — or ask the Assistant in plain language.
- Two ways to run it
- Hosted — we operate it. Self-Hosted — isolated in your subscription. Same Teams tab, same widget, Azure AI Foundry you own.
One platform, a widget for every domain
Give each website, brand or app its own branded widget — routing, sign-in, files and hours per widget — all from one Teams tab.
Self-Hosted today: Today the Self-Hosted runtime ships one website widget (Growth+) on the origins you set at deploy time, answered by the same Teams tab.
Per-widget controls
Change any of it from the dashboard; no redeploy.
- Look and every visible string
- Allowed origins
- The Foundry agent that answers first
- Sign-in mode
- File rules and permissions
- Business hours
- Inactivity timeout
- Fallback contact
Each widget carries a list of allowed origins, and that list drives the Content-Security-Policy frame-ancestors header on the widget host, so a browser refuses to render the widget on a domain you did not list.
One script tag on the page. Which widget answers, and which domains may render it, follow from the widget you point that tag at.
The widget is mobile-first: it respects the host page's safe areas, follows the on-screen keyboard as it opens and closes, and behaves with the browser back button on Android.
One example per domain
shop.example.com
AI answers product and delivery questions first; a sales rep in Teams takes over when the visitor taps “Chat with a Human”. Anonymous sign-in.
Included with Point of Contact — Hosted. Self-Hosted today: one AI-first widget, claimed and closed in the Teams tab.
The AI is yours. The people are in Teams. The conversation moves between them.
Your Foundry. Your tokens. Our platform.
Provisions an Azure AI Foundry resource and project into your subscription, or connects yours. One assistant, a prompt you edit; Microsoft bills inference.
Included with Point of Contact — Hosted
Your agents. Your tokens. Our platform.
Bring the Foundry agents you own — model, instructions, knowledge. Microsoft bills your inference; we never run the model or override your instructions.
Coming to Self-Hosted
Model choice, a registry of several Foundry agents with routing, and a summary handed to the receiving agent.
- No cap on AI conversations and no per-resolution fee on any plan; your single variable cost is inference on your Foundry, billed by Microsoft at Microsoft's rates.
- The AI never trains on your data, on any product: inference runs on Azure AI Foundry you own, under your own Azure agreement, and we never run the model.
AI first or human first. Hand off as often as it takes.
Start with AI or a Teams user; "Chat with a Human" brings your team in; transfer, hand back, reopen — one thread throughout.
Self-Hosted today: Today on Self-Hosted the AI answers first; your team sees the live chats in the Teams tab, takes over a chat the AI is answering and closes it there.
An unanswered request for a person returns to the AI after 5 minutes, an unaccepted transfer returns to the original agent after 2, and a chat with no activity closes itself after 10 minutes.
Included with Point of Contact — Hosted
How a handoff runs
- 1 The visitor asks; the Foundry agent you registered answers first — or a Teams user does, if you start human-first.
- 2 The visitor taps “Chat with a Human” — or asks in their own words — or the agent decides a person should take it; the right Teams users are notified, filtered by live presence.
- 3 An agent claims the chat with the whole transcript in view. If nobody claims it within 5 minutes, the chat returns to the AI with a note to the visitor.
- 4 Transfer to a colleague or a group, take over a colleague’s chat, hand back to the AI, close — and reopen a closed chat when the visitor returns. Idle chats close after 10 minutes.
Hours, presence, and always another point of contact
Hours decide when a human can be offered; live presence decides who is notified; outside both, the widget shows your fallback contact.
No offline form, no ticket queue: the fallback contact is the path, and the AI keeps answering.
See how a handoff runsAnonymous by default. Registered when it matters.
Visitors register and sign in inside the widget, through your own Entra External ID tenant and the identity providers you set there.
Per widget you choose anonymous, optional or required sign-in. Registration and profile happen in the widget, against an External ID tenant you own; the service validates a real OIDC token. Signed-in users open their Past conversations, download a PDF transcript and unlock the file space. Admins do not manage external users in the app: the directory is yours; the widget is the front door.
Sign-in mode is a per-widget setting: anonymous, optional or required. A support portal can require it while the marketing site never mentions it.
The service validates the token's signature, issuer, audience, tenant and object id before a session becomes an identified one; we store the object id, never credentials.
Files where necessary — a secure Azure space, for review
Registered users upload, draft and submit files into a private Azure Blob space your team reviews from the same conversation.
- 15 MB per file by default, 100 files and 500 MB per registered user, with the allowed file types set per widget; submitted files are kept in a 28-day backup.
- Draft, submit for review, request deletion. Group membership and widget admins decide who sees what; files appear as bubbles in the thread.
Where a registered user can upload a file, it goes to a private Azure Blob space for your team to review. The AI never reads it.
See what people actually contact you about
An AI pass over closed conversations — by date, agent and domain — surfaces top topics and suggested improvements for AI and team.
Topic Analysis runs across your closed chats for the range you pick, groups them by topic, shows the domains behind each, and writes per-topic recommendations for your human agents and your AI agent. The closed-chat report adds source domain, agents, transfers and handling time. For continuous improvement: no sentiment scores, no trend charts, no per-user profiles.
Filter by date, agent and widget domain; each run classifies up to 200 new closed chats and reuses the classifications it already has, so a large backlog is worked through in predictable, priced steps.
A read-only view of closed conversations, filtered by date, agent type, specific agent and source domain: chats resolved, time to resolution, human handling time, transfers per chat and who closed each one.
Self-Hosted today
The cost dashboard shows token usage and its estimated cost; the median projection is a Professional entitlement, and infrastructure spend stays in Azure Cost Management.
Chat that lives where your team already works.
Agents and admins, one Teams tab
Agents and admins work in one Teams personal tab — take, transfer, close, review files, run analysis, change settings — with Teams SSO and activity-feed alerts.
Self-Hosted today: Today the Self-Hosted Teams tab lists the live chats, takes over a chat the AI is answering, replies and closes it — with Teams SSO — from an app package built for your tenant.
Admins in the same tab
Tell the Assistant what you want — "create a widget for acme.com" — and it changes the settings through the app's own endpoints, as you.
Not a notification bridge
Helpdesk suites and Teams "bridges" push a notification into Teams; the work still happens in their console, on their tokens. Here the queue and the live chat are the Teams tab today; settings, transfers, files and analysis in the same tab — Coming to Self-Hosted
Point of Contact 3 live chats Blue, next-day delivery
Damaged order, replacement
Quote documents submitted
None of the 14 chat and contact-centre products we compared in August 2026 combines all three: an agent workspace inside a Microsoft Teams tab, a website widget per domain, and hand-off from AI to a person and back again. Among the products that do put agents in Teams, none documents a way to hand the conversation back to the AI. Point of Contact — Self-Hosted ships the Teams workspace today; per-domain widgets and the round trip are Hosted capabilities — Coming to Self-Hosted. The comparison and its sources are published; tell us if we missed one and we will update it.
Two ways to run it
Hosted — we operate it. Self-Hosted — isolated in your subscription. Same Teams tab, same widget, Azure AI Foundry you own.
| Point of Contact — Hosted | Point of Contact — Self-Hosted | Self-Hosted — Private plan | |
|---|---|---|---|
| Who operates it | We do. | You do — a managed application in your subscription; we publish releases, you apply them. | You do — a managed application in your subscription with no publisher role at all. |
| Where the conversation record lives | Our Azure. | Your Azure. | Your Azure. |
| Publisher access to it | Yes — we operate it. | JIT only — you approve, time-boxed to 8 hours, logged (Entra ID P2). | None — no publisher access at all. |
| The AI | The Azure AI Foundry project you own — your agents, your model; Microsoft bills your inference. | A Foundry resource and project provisioned into your subscription (gpt-5-mini deployment) or a project you connect; Microsoft bills your inference. | A Foundry resource and project provisioned into your subscription (gpt-5-mini deployment) or a project you connect; Microsoft bills your inference. |
| Availability today | The full platform, today. | The shipped core today — one widget (Growth+), AI first, claim and close in Teams, cost dashboard, signed updates; rows marked Coming to Self-Hosted are on their way to the same isolated runtime. | The shipped core today — one widget, AI first, claim and close in Teams, cost dashboard, signed updates, and every capability of the Professional tier; rows marked Coming to Self-Hosted are on their way to the same isolated runtime. |
| Updates | We push updates. | You initiate; signature, version chain and SHA-256 verified before apply. | You initiate; signature, version chain and SHA-256 verified before apply. |
| Price | Start $49 / 3 seats (+$20) · Growth $99 / 7 (+$18) · Professional $199 / 15 (+$15), plus your Foundry inference — no estimate published. | Same ladder, tier derived from the seats you enter, plus your Azure at cost (about $3–15 / $10–30 / $60–120 a month by tier) and inference. | Negotiated flat fee plus your Azure at cost. |
| Procurement and posture | Azure Marketplace SaaS (Coming to the Azure Marketplace). The honesty block applies; we are the processor for the conversation record. | Azure Marketplace managed application (Coming to the Azure Marketplace); can count toward an Azure consumption commitment; Entra ID P2 mandatory. | Private plan by arrangement; no Entra ID P2 needed. |
| Best for | You need the whole platform now and are fine with us storing the record. | Security- and residency-led buyers who want the stack in their subscription and can approve a time-boxed elevation. | Reviews that will not accept even a consent gate. |
On Hosted: model processing is yours; conversation storage is ours. Your Foundry project handles every prompt and completion under your own Azure agreement — we never see them. The conversation record — transcripts, files, session state — is stored in our Azure, and our operations team can access it to support you. If your review requires that no vendor can reach that record, that is exactly what Self-Hosted and Private are for.
Ask two questions
Ask two questions. Do you need every capability today? Hosted has the full platform now; Self-Hosted is receiving it (see the availability table). Does your review require that no vendor can reach the conversation record? That is Self-Hosted — or Private if a time-boxed elevation you approve is still too much. Both run the same Teams tab and the same widget, on Azure AI Foundry you own.
Two rows do the deciding: where the conversation record lives, and who can reach it. Everything else is identical or operational.
Moving between Hosted and Self-Hosted is a change of edition, not a change of product: same Teams app, same seat ladder, same widget. Agent configuration does not carry across — you set the system prompt again on the new deployment.
Compared with 14 products (August 2026)
None of the 14 chat and contact-centre products we compared in August 2026 combines an agent workspace inside a Microsoft Teams tab, a website widget per domain, and hand-off from AI to a person and back again. Per-domain widgets and that round trip are Hosted capabilities today — Coming to Self-Hosted.
| Vendor | Category | Agent workspace in Teams? | Widget per domain | Whose model, whose tokens | Handoff modes | Pricing model (08/2026) |
|---|---|---|---|---|---|---|
| Point of Contact — Hosted | Teams-native chat platform | Yes — One Teams tab | Yes — per widget | The Azure AI Foundry project you own — your agents, your model; Microsoft bills your inference. | AI to human and back, transfer, takeover, hand-back, reopen | Start $49 / 3 seats (+$20) · Growth $99 / 7 (+$18) · Professional $199 / 15 (+$15), plus your Foundry inference — no estimate published. |
| Point of Contact — Self-Hosted | Teams-native chat platform, managed app in your subscription | Yes — Teams tab today | One widget today | A Foundry resource and project provisioned into your subscription (gpt-5-mini deployment) or a project you connect; Microsoft bills your inference. | AI first, claim and close today | Same ladder, tier derived from the seats you enter, plus your Azure at cost (about $3–15 / $10–30 / $60–120 a month by tier) and inference. |
| Microsoft Teams Live Chat / Customer Connect (retiring) | First-party Microsoft 365 Business website live chat relayed into Team… | Auto-assignment, not claim: Teams assigns the request to the next available support member based… | not documented (checked 2026-08-16) — docs describe a single tenant-level widget bound to one tr… | AI usage is billed as Copilot Studio Copilot Credits (tenant packs… | not documented (checked 2026-08-16) | Included in per-user M365 Business licence (no separate charge… |
| Microsoft — Dynamics 365 Contact Center + Copilot Studio | Enterprise CCaaS (omnichannel contact center) + low-code AI agent buil… | No — representatives accept/handle chats in Copilot Service workspace (Dynamics 365 web app)… | Yes — many chat channels/workstreams per org; rulesets group rules 'if you support multiple bran… | Partial — with a bring-your-own Foundry model in a prompt tool… | Yes — representative can transfer the conversation to a (second) AI agent in the same or another… | Per representative seat (user/month, paid yearly) + consumption Copilot Credits for AI agents (2… |
| Social Intents | Live chat + AI chatbot SaaS answered from Teams/Slack/Google Chat/Zoom… | Yes — invitation posted in Teams channel; any member clicks Join… | Yes, plan-capped: Lite 1 / Starter 1 / Basic 2 / Pro 5 / Business 10 (Enterprise 30+)… | Yes with BYOK — provider bills the customer's OpenAI/Anthropic/Google account… | No — docs: 'There is no automatic way to transfer the conversation back to the chatbot'… | Flat per plan (monthly/annual, 'Save 20%' annual) with conversation caps… |
| Intercom (Fin AI Agent) | AI-first customer service suite (helpdesk + Messenger + Fin AI Agent)… | No — no native Teams integration; agents claim/reply in Intercom Inbox (web/desktop/mobile)… | Yes on Expert plan only (or legacy Scale) — up to 300 brands per workspace… | No — usage billed by Intercom per outcome, not by customer's cloud tokens | Partial — 'It's not possible to reassign a conversation to Fin from a teammate inbox'… | Per seat (Essential $29 / Advanced $85 / Expert $132 per seat/mo… |
Four of fourteen products, cells trimmed. As of August 2026, list prices, subject to change — verify before publishing. The other ten, the full text and every source are on the comparison page: Zendesk, Salesforce Agentforce, Freshworks, Azure Communication Services, Tidio, Crisp, Zoho SalesIQ, HubSpot, LiveChat and Chatwoot.
Private plan — Negotiated flat fee plus your Azure at cost.
The leading AI helpdesks bill about a dollar per AI resolution on their model. Point of Contact bills per human seat; the AI runs on Foundry you own, and Microsoft bills the tokens at Microsoft's rates.
Microsoft is retiring Live Chat for Teams — new setups closed 6 August 2026; service ends 5 October 2026. Point of Contact keeps the website widget answered by people in a Teams tab and adds AI first on a Foundry assistant you own — more than one domain, registration and files are coming to Self-Hosted.
One plan, priced on the seats you enter
No per-resolution fee, no per-conversation charge: the licence is per human seat, and inference is Microsoft's line at Microsoft's rates.
| Seats | 1–3 | 4 | 5 | 6 | 7 | 10 | 12 | 13–15 | 20 |
|---|---|---|---|---|---|---|---|---|---|
| $ / month | 49 | 69 | 89 | 99 | 99 | 153 | 189 | 199 | 274 |
| Tier | Start | Start | Start | Growth | Growth | Growth | Growth | Professional | Professional |
The entitlement tier follows the number you enter: Start up to 5 seats, Growth 6–12, Professional 13 and above — and you always pay whichever band is cheapest for that count.
Point of Contact — Self-Hosted
- One plan, one ladder: 3 seats $49 · 7 seats $99 · 15 seats $199 per month; extra seats $20 / $18 / $15 by tier — e.g. 20 seats = $274. Plus your Azure at cost.
- Two-part cost: a flat seat-ladder licence to us, and your own Azure at cost — infrastructure and inference billed to you directly by Microsoft at Azure's published rates. No vendor mark-up.
- Your Azure runs at cost: roughly $3–15 a month at Start scale, $10–30 at Growth, $60–120 at Professional, plus inference — at Azure list prices; Azure Cost Management shows the actuals.
- One plan, priced on the seats you enter: the tier follows your seat count on the published ladder, the seat guard meters usage, and growing never means a new deployment.
- The stack is serverless in your subscription — Flex Consumption Functions and serverless Cosmos DB scale to zero when idle and pay per request; Professional defaults to one warm instance.
- An Azure Marketplace managed-application purchase can count toward an Azure consumption commitment.
Self-Hosted (public plan)
AI + human chat for Teams, deployed into your own Azure. Pay for the seats you enter.
Private (by arrangement)
Negotiated plan with no publisher role on your deployment at all. By arrangement.
The Private plan is a negotiated flat fee — no seat metering, no meters at all — plus your Azure at cost, billed by Microsoft.
Point of Contact — Hosted
Three plans, one seat ladder
- Start $49 for 3 seats (+$20 per extra seat) · Growth $99 for 7 (+$18) · Professional $199 for 15 (+$15), per month, plus your Foundry inference.
- Two line items: the plan price to us — from $49 a month — and inference on the Azure AI Foundry project you own, billed to you by Microsoft. We run the infrastructure.
- We publish no inference estimate: you choose the model and the volume on the Foundry project you own, and Microsoft bills you at Microsoft's rates.
- An Azure AI Foundry project you own is required on every Hosted plan; there is no plan on which we supply the model.
- Add agents as you grow: extra seats are metered per seat on your plan, and moving from Start to Growth to Professional changes the licence, not the deployment.
Included with Point of Contact — Hosted
Start
3 seats
3 agent seats and unlimited AI triage, on an Azure AI Foundry project you own.
Growth
7 seats
7 seats, the embedded website widget, and priority email support. Your model, always.
Professional
15 seats
15 seats, custom branding, Topic Analysis over closed chats, and 24/7 support with a named CSM.
Coming to the Azure Marketplace — commercial terms live on the listing.
Every domain that needs external chat
Same platform, same Teams tab — a different front door on every website you run.
Self-Hosted today: Today on Self-Hosted the AI answers first; your team sees the live chats in the Teams tab, takes over a chat the AI is answering and closes it there.
shop.example.com
B2C sales
A shop puts the widget on its storefront. The Foundry agent answers sizing, stock and delivery questions from the knowledge you gave it; when a visitor wants a person, a rep in Teams claims the chat with the whole thread in view. Anonymous by default.
Will the blue one ship before Friday?
Yes — it is in stock and goes next-day to your postcode. Want me to hold one while you decide?
Frequently asked questions
What is Point of Contact?
Hosted or Self-Hosted — which one?
Do our agents need to learn a new tool?
How does AI-to-human handoff work?
What happens when nobody is online?
Can we run more than one website or brand on it?
Whose AI model is it, and who pays for it?
Can we use several AI agents?
How does visitor sign-in work?
Does the AI read uploaded files?
Do you use our data to train AI models?
How does pricing work?
Does Self-Hosted need Microsoft Entra ID P2?
How do updates reach a Self-Hosted deployment?
Moving from Teams Live Chat?
What does deployment involve?
Is there a free trial?
How do I contact the team?
Still have questions?
Get in touchTalk to us
A 20-minute call: your domains, your Teams setup, your Foundry — and whether Hosted or Self-Hosted fits your review.
Email us directly
info@innovations24.comResponse time
We reply within one business day.
Status
Coming to the Azure Marketplace