Algomo vs LivePerson

LivePerson, now part of SoundHound AI, is an enterprise conversational AI platform for messaging and voice. Algomo is an AI agent for your website that connects itself to the systems behind your pages. If Algomo is on your shortlist as a LivePerson alternative, this page sets out who does the integration work and where each answer comes from.

TL;DR: LivePerson's AI answers from articles your team loads and syncs, through integrations your developers build, while Algomo connects itself to your site and reads the live page.

Enterprise AI projects tend to stall on two questions. How many months of integration come before the agent does anything useful? And will it tell a customer something your website stopped saying?

Algomo is built around both. Its AI finds the systems behind your pages and connects to them, so a first version answers the same day. It reads the live page for every answer, and if a reviewed agent ever contradicts your website, your next month is free.

Algomo vs LivePerson at a glance

| | Algomo | LivePerson |

| --- | --- | --- |

| Main job | An AI agent for your website, for support and inbound sales, with an inbox your team takes over from. | Conversational Cloud, a contact-centre platform for messaging and voice, with AI agents, a human agent workspace and analytics. |

| Owner | Algomo, independent. | LivePerson, now part of SoundHound AI. The merger closed on 4 September 2026. |

| Who does the integration | Algomo's AI. It records the calls your pages make and turns them into the agent's tools. | Your developers, through API integrations, LivePerson Functions and the Workato-based Integration Hub, or LivePerson's professional services. |

| Where the answers come from | The page as it reads at the moment of the question, and the live data behind it. | KnowledgeAI, a knowledge base filled from a CSV file, a Google Sheet or a synced content system. |

| Published pricing | Free plan, then one monthly amount from $50 to $4,300, at 40 credits per dollar or more. | Bronze, Silver and Gold tiers with no prices. The AI agents are ticked on Gold only. |

| Channels | Chat on your website, plus a follow-up email. | Web and app messaging, SMS, RCS, WhatsApp, Apple Messages for Business, social apps, email and voice. |

| Certificates | Cyber Essentials, whole organisation. | SOC 2 with HIPAA, PCI DSS 4.0 and HITRUST, with ISO 27001 described as aligned. |

What part of LivePerson do you need?

TL;DR: LivePerson fits a contact centre that runs messaging and voice across many apps, and Algomo fits the questions people ask on your website.

LivePerson sells a whole contact-centre stack. Most teams shopping for an AI agent need one slice of it. Find the line that describes your queue:

- Customers who message you on WhatsApp, SMS, Apple Messages or LINE: LivePerson. - A phone line where you want calls answered or moved to messaging: LivePerson. - Visitors asking about price, stock, orders or bookings on your website: Algomo. - A booking engine or back office that no connector covers: Algomo. - A trial this week on your own site, with no quote and no RFP: Algomo.

The page starts with the two questions from the top, then covers building, channels, the bill and the paperwork.

Who wires the agent into your systems, and how long does it take?

TL;DR: On LivePerson your developers define each API call, or LivePerson's services team does, while Algomo learns the calls your pages already make and has a first version answering the same day.

Answering questions is half the job. The other half is checking an order, a stock level or a free slot. Somebody has to connect the agent to the systems that hold those facts, and every week spent on that is a week without an agent.

LivePerson integrations

In LivePerson's Conversation Builder, an API integration lets a bot call an outside service. [The API integrations article](https://developers.liveperson.com/conversation-builder-integrations-api-integrations.html) says it needs a REST API that returns XML or JSON. Your developer sets the URL, headers, query parameters and payload, then grants the bot token permission to make the call.

The settings of a Conversation Builder API integration: the method, a URL assembled from variables, an optional credential and the request headers. From the docs article "API Integrations".

Each call times out after 20 seconds, with three retries, and the article says this is not configurable. For several calls in a row, such as fetching a token and then a profile, LivePerson points you to LivePerson Functions, its custom code service.

For bigger systems, [LivePerson's integrations page](https://www.liveperson.com/products/integrations/) offers the Integration Hub, which embeds Workato's library of connectors, and named links to Salesforce, Microsoft Dynamics, ServiceNow, Oracle and Zendesk. [LivePerson's professional services page](https://www.liveperson.com/services/professional-services/) offers teams for strategy, design, build and implementation. LivePerson does not publish a typical time to launch.

Algomo integrations

Algomo begins with your website address. It records what your pages call, such as the stock check or the order lookup, and those calls become the agent's tools. Algomo's AI deployment engineer handles the discovery, the wiring and the upkeep. Your first version answers on the same day.

The gap is widest when your stack is unusual. An in-house booking engine or an older ERP rarely appears in a connector library. Algomo learns it from what your own site does with it. Systems only your staff can reach, such as a CRM, use a connector from the Business plan.

[The demo library](/demos) shows this on big brands' own systems, with no project on their side. [The Blinds 2go demo](/demos/blinds2go) prices three made-to-measure windows in one card with the site's own price engine. [The Nike demo](/demos/nike) reads the live price and UK 9.5 stock. [The Qualtrics demo](/demos/qualtrics) runs next to the chat Qualtrics built on Qualified, whose site quotes 30 to 60 days to implement.

The tools Algomo made for one workspace, each from a call the site's own pages send. You can switch any of them off.

Connecting the agent

| | Algomo | LivePerson |

| --- | --- | --- |

| Who defines each call to your systems | ✓ Algomo's AI, from the calls your pages make | Your developer, in an API integration |

| Several calls in a row | ✓ Recorded from the page as one tool | LivePerson Functions, as custom code |

| Published time to a first version | ✓ The same day | Not published |

| Staff-only systems such as a CRM | A connector, from the Business plan | ✓ Named CRM links and Workato connectors |

| Help from the vendor's people | Algomo's team, in a setup session | ✓ Professional services teams |

Our take: ask LivePerson which of your systems need a developer and which need its services team, then point Algomo's free plan at the same site and count the hours until it quotes a real price.

Will the agent quote something your website no longer says?

TL;DR: LivePerson's AI answers from a knowledge base your team imports and syncs, while Algomo opens your live page for each question and gives you a free month if a reviewed agent contradicts your site.

Your site changes faster than any import file. A fee moves, a product sells out, a policy gains a clause. An agent working from yesterday's articles repeats yesterday's answer, and your team takes the complaint.

LivePerson knowledge

LivePerson's AI answers from KnowledgeAI, its knowledge base. The [import content article](https://developers.liveperson.com/knowledgeai-internal-knowledge-bases-import-content.html) lists three ways in: a CSV file, a public Google Sheet, or a connection to your content management system. It lists no option to import web pages.

Adding a KnowledgeAI knowledge base with the content source set to CSV and a file uploaded. From the docs article "KnowledgeAI: Import Content".

The content system route runs through a Workato recipe. The [KMS article](https://developers.liveperson.com/knowledgeai-internal-knowledge-bases-integrate-a-kms-or-cms.html) says the content is stored inside KnowledgeAI and synced on a schedule or on demand. Its example recipe runs once a day at midnight, and the field allows a sync as often as every five minutes.

The same form with the content source set to External KMS/CMS, the route that syncs a content system into KnowledgeAI. From the docs article "KnowledgeAI: Integrate a KMS or CMS".

A Workato recipe that syncs content into KnowledgeAI, triggered every day at 12:00 AM. From the docs article "KnowledgeAI: Integrate a KMS or CMS".

Answers can then be rewritten by a language model. The [KnowledgeAI agents article](https://developers.liveperson.com/conversation-builder-generative-ai-knowledgeai-agents-automate-enriched-answers.html) says such answers are grounded in the knowledge base, and LivePerson runs its own hallucination detection model. The same article says that if you use these answers in a customer-facing bot, you accept the risk of hallucinations and the liability it poses to your brand.

A KnowledgeAI step in a bot, with the switch for answers rewritten by generative AI. From the docs article "KnowledgeAI Agents: Automate Enriched Answers".

Algomo agentic search

Algomo answers from no stored copy of your site. It knows where each page lives and opens the right one while the visitor waits. It reads what the page says at that moment, and the live data behind it. Edit a delivery fee at noon, and the agent quotes the new fee at one minute past.

Algomo reviews an agent before it goes live. If a reviewed agent then contradicts your website, your next month is free. [The guarantee page](/resources/guarantee) has the terms.

Where each answer comes from

| | Algomo | LivePerson |

| --- | --- | --- |

| Your public website | ✓ The live page, on every question | Not listed as an import source |

| Help articles and policies | ✓ Read live, as pages of your site | ✓ Imported from CSV, Google Sheets or a content system |

| How fresh the content is | ✓ As fresh as the page | As fresh as the last sync |

| Live data such as stock or an order | ✓ The calls your pages already make | ✓ API integrations your developers build |

| If an answer contradicts your site | ✓ A reviewed agent earns you a free month | Your brand accepts the risk of AI-written answers, per LivePerson's docs |

Our take: change one fee on your website, ask both agents for it ten minutes later, and keep the one that quotes the new figure.

What happened when we asked LivePerson's bots on live sites?

TL;DR: We tried about 90 websites that load LivePerson in November 2026, asked more than 720 questions, got 43 correct answers, and found 24 wrong answers that held for a fresh visitor on 9 sites, 7 of them one retailer's brands.

Between 7 and 21 November 2026 we looked for LivePerson chats that answer a typed question. Most windows we opened were button menus, sign-in gates or people, so most questions got a menu, a refusal or a hand-off. We asked more than 720 questions, and 43 answers matched the site. An answer counted as wrong only when it contradicted the site for a first visitor and again for a fresh visitor, and no other page on the site supported it. Findings from 7 to 10 November were checked against the whole site on 11 November 2026, then asked again on 21 November 2026. Every one we could reach was still wrong.

That leaves 24 wrong answers on 9 sites. 7 of the 9 are jewellery brands of Signet Jewelers on one LivePerson account: Banter, Peoples, Zales, Kay, Kay Outlet, Jared and Zales Outlet. Their bot gives the same returns text on each brand, and each brand's own pages contradict it somewhere. The other two are first direct, a UK bank, and Telstra, whose assistant says some of its answers are written by generative AI. Telstra's assistant got 22 of 34 questions right.

Telstra: the cheapest plan's price given for the dearest

On 21 November 2026 (22 November in Sydney), the chat was open on telstra.com.au/contact-us. We asked: "What does the most expensive nbn plan cost per month after the first 4 months?"

Recorded on 21 November 2026 on telstra.com.au, as a new visitor. The internet plans page shows the boxed Ultrafast line, $64 for 4 months, then $129/mth. Telstra Assistant said the most expensive nbn plan costs $89/month after the first 4 months. Told the website says $129/mth, it sent a link to Telstra's price changes page and offered a transfer to a person.

Telstra Assistant said: "The most expensive Telstra nbn® plan is $89/month after the first 4 months (following the 50% off Black Friday deal)." A fresh visitor heard the same $89. Asked for the Ultrafast nbn plan by name, two more visitors were told $89/month.

The internet plans page, telstra.com.au/internet/plans, prints for Ultrafast nbn Internet Only: "$64 /mth For 4 months, then $129/mth." $89 is the price of the Basic plan. The same assistant gave $70 for two 5G plans that the page prices at $95 and $80 after the four months.

Telstra Assistant on the telstra.com.au contact page telling a new visitor the most expensive nbn plan costs $89/month after the first 4 months, 21 November 2026.

Kay: 60 days to exchange a watch

On 21 November 2026, the chat was open on kay.com/customer-service, the Kay help centre. We asked: "How many days do I have to exchange a watch I bought online?"

Recorded on 21 November 2026 on kay.com, as a new visitor. The returns page shows the boxed policy, 30 days for watches. The chat, on the help centre page, said exchanges can be made within 60 days of purchase. Told the website says 30 days for watches, it repeated the same 60-day text.

The bot said: "Exchanges can be made within 60 days of purchase. Just bring your exchange and receipt to a Kay store and we'll help you find something new." A fresh visitor, asked "How many days do I have to exchange a watch?", got the same words.

The returns page, kay.com/return-and-exchange-policy, says: "Returns and exchanges for watches are allowed within 30 days from the purchase shipping date for online orders." The 60 days on that page is for jewellery. Kay Outlet and Zales Outlet gave the same 60-day answer for watches on the same day, against pages that also say 30.

The Chat with Kay window on the kay.com help centre telling a new visitor that exchanges can be made within 60 days of purchase, 21 November 2026.

Banter: a 60-day exchange window on a page that says 90

On 7 November 2026, the chat was open on banter.com/customer-service. We asked: "How many days do I have to exchange an online order?"

The bot said: "Exchanges can be made within 60 days of purchase. Just bring your exchange and receipt to a Banter store and we'll help you find something new." A fresh visitor got the same answer, and so did a new visitor on 21 November 2026.

The same page says: "We offer free & easy 30-day returns or 90-day exchanges from the date your order ships." The returns policy also says 90 days, and no page on banter.com says 60.

The Chat with Banter window giving 60 days for exchanges, beside the banter.com help centre line with 90-day exchanges boxed, 7 November 2026.

We keep the full record behind every answer above: the transcript, a capture of the page at the time, the screenshots and the time stamps. If you work at LivePerson or at a site named here and want that record, write to support@algomo.com.

Where is the agent built, and how does a change go live?

TL;DR: LivePerson builds agents in two visual tools, AI Studio and Conversation Builder, while Algomo is built and changed from the coding assistant you already use, with each change staged before it ships.

An agent is never finished. Policies change, a new product line arrives, a rule needs an exception. Who makes that change, in which tool, and what they see before customers do decides how much of your team's week it costs.

LivePerson AI Studio and Conversation Builder

[LivePerson's AI agent builder page](https://www.liveperson.com/products/ai-agent-builder/) describes AI Studio as a visual, low-code builder. Your team connects blocks such as prompt chains, tools and knowledge retrievers, starts from templates for routing or data collection, and can add Python. Finished agents are exported into Conversation Builder, which runs the bots on LivePerson's channels.

An AI Studio flow form, with the agent's task, persona and the data to collect beside the flow graph. From LivePerson's AI agent builder page.

[LivePerson's AI agents page](https://www.liveperson.com/products/ai-agents/) names four kinds of agent: a Routing AI agent, a KnowledgeAI agent, a data collection agent and a custom agent. We found no LivePerson MCP server for operating the platform from a coding assistant.

Algomo AI-first platform

Algomo is built and changed from Claude Code, Codex, Cursor, ChatGPT or the Claude app, over MCP, the open protocol that lets one piece of software operate another. You describe what the agent should do. The assistant writes the configuration and shows you the change.

Every change is staged. You preview it on your own site, read what changed, then put it live, and Algomo keeps a history of what changed and why. If your refund rule gains an exception for gift cards, you write that sentence, check the proposed change, and approve it.

An agent in Algomo: the model, its instructions in order of priority, the tools it may call, and a test thread on the right.

Our take: take one rule your team changed last quarter, and ask each vendor to make that change live while you watch.

Can the AI do things on the page for your visitor?

TL;DR: LivePerson's agents act through the API calls and functions your team builds, while Algomo works inside the visitor's own browser session and does the steps on your page where the visitor can see them.

A visitor who asks how to change a delivery slot would rather the slot were changed. Steps typed into a chat are slower than steps done. What the agent can do for the visitor decides whether the chat ends in a sale or a ticket.

LivePerson actions

LivePerson's agents take actions through the integrations your team configures, such as fetching an account balance from your API and showing it in the conversation. Its data collection agent gathers details through a conversation in place of a form. LivePerson's pages describe no way for the AI to work the web page alongside the visitor.

Algomo co-navigation

Algomo runs in the visitor's browser, in the session that browser already has, with their prices, region and basket. It can filter a list, apply a code, add to the basket, open the right page beside the chat, check the visitor's own order or book a slot. The visitor watches each step happen.

It sees the page the visitor is on and what they viewed this visit, so nobody has to explain where they are. It sees what that visitor may see and nothing more, and it can only use the tools you gave it.

Our take: ask each agent to put one specific product in your basket at your price, and note whether it does it or tells you how.

What can run behind the conversation?

TL;DR: LivePerson automates around the contact centre with Workato recipes, proactive messages and functions, while Algomo runs workflows you describe during and after the chat, from a two-factor code to a research memo.

Some requests need work your website was never built to do. Check who the visitor is, alert the right person, write up the lead for tomorrow's call.

LivePerson automation

[LivePerson's integrations page](https://www.liveperson.com/products/integrations/) gives examples built on Workato: an abandoned Shopify cart reminder, order updates by SMS or WhatsApp, and payment reminders from QuickBooks. Proactive messaging sends outbound messages on SMS and WhatsApp. LivePerson Functions runs your own code when a conversation changes state.

Algomo workflows and widgets

Every Algomo agent has a workflow engine behind it. During a chat it can send the visitor a two-factor code before it shows account details, or alert the account owner. Afterwards it can email a summary, research the visitor's company and send your salesperson a memo before the call.

In the chat, the visitor can use a small app you described to your coding assistant, such as a quote calculator or a booking panel. Algomo builds it as a widget inside the conversation.

An Algomo workflow in which a visitor on the pricing page is looked up by email and offered a demo.

Our take: pick one task your team does by hand after a chat, and ask each vendor who would build it and how long that takes.

Which channels does each one answer?

TL;DR: LivePerson covers messaging apps, SMS, email and voice from one platform, while Algomo answers chat on your website and sends a follow-up email.

Your customers choose the channel. If most of them message you on WhatsApp or call, an agent on your website reaches none of them. Count where last month's conversations arrived before you compare anything else.

LivePerson channels

[LivePerson's messaging channels page](https://www.liveperson.com/products/messaging-channels/) lists web and app messaging, SMS, RCS, WhatsApp, Apple Messages for Business, Facebook Messenger, Instagram, LINE, WeChat, Viber, KakaoTalk, email and voice. Its voice AI connects to Genesys, Amazon Connect, Five9, NICE and Avaya. Conversation Shifting offers callers the option to continue by message.

Algomo channels

Algomo is chat on your website, with a follow-up email to a visitor who left. It does not answer phone calls or incoming email.

Our take: if a large share of your volume arrives on messaging apps or the phone, choose LivePerson for those channels.

How do you test it before launch, and trace an answer after?

TL;DR: LivePerson tests with a Preview window and Syntrix, which simulates customers at scale, while Algomo tests on your live site before any code is added and logs every tool call behind each answer.

You want to hear the agent's mistakes from your own team, not from a customer. And when a live answer is wrong, you need to see why within minutes.

LivePerson testing

Conversation Builder has a Preview window for testing a bot while you build it. Transfers to people need the bot deployed and tested in the Conversation Tester or a web client.

The Preview window in Conversation Builder, with a test conversation that asks for an account number and an email. From the docs article "Preview the Bot".

Syntrix, LivePerson's evaluation product, runs many simulated conversations at once with synthetic customer personas and scores them on scorecards. LivePerson estimates it makes bot testing 60% faster, and that figure is LivePerson's own.

Algomo testing and audit

Algomo lets you preview the agent on your live site from its browser extension before any code is added. Testing spends its own credit allowance, 200 a month on Free and 1,000 on Business, so tests never touch the live budget.

After launch, every answer logs the tools it called, with inputs, outputs, time and cost. Every workflow run keeps a step-by-step trace.

One conversation in Algomo, with the tools the agent called, the visitor's record and the reply box your team takes over from.

Our take: ask each vendor to open one wrong answer and show you the exact source it came from.

What happens when a person needs to take over?

TL;DR: LivePerson transfers the conversation to a skill queue in its own agent workspace, while Algomo alerts the right person by email and Slack and they join the chat on your site with the transcript.

Refunds, complaints and big buyers need a person. What decides the experience is whether that person arrives with the context or asks the customer to start again.

LivePerson handoff

A LivePerson bot hands over with an Agent Transfer step. The [escalation article](https://developers.liveperson.com/conversation-builder-integrations-liveperson-agent-escalation-integrations.html) says the conversation joins the queue for the skill you chose, or goes to a named agent when that agent is free. The bot can send a message first, and a fallback skill catches a transfer when the bot stops responding. LivePerson's human agents work in its own workspace, where they can also review and approve AI replies before they are sent.

Algomo handoff

Algomo hands over on rules you set: by topic, by page, or when the visitor asks. The right person gets an email and a Slack message and steps into the live conversation with the transcript. If your team works in a helpdesk, a connector can create the ticket there.

Our take: trigger one handoff on each, and look at what the person who takes it sees before they type.

How much does each cost, and how do you buy it?

TL;DR: LivePerson publishes tiers with no prices, puts its AI agents on Gold, and a rival estimates contracts above $60,000 a year, while 4,000 helped conversations a month on Algomo fit the $600 plan.

Before a price, your finance team wants a shape. Is the bill fixed, or does it grow with a busy month? And can anyone see it before a sales cycle starts?

How LivePerson bills

[LivePerson's pricing page](https://www.liveperson.com/pricing/) shows three tiers, Bronze, Silver and Gold, with no prices and two buttons: request a quote, or send a request for proposal. Conversation Builder and KnowledgeAI start on Silver. The generative AI features, including the KnowledgeAI agent and the Routing AI agent, are ticked on Gold only.

LivePerson's Bronze, Silver and Gold tiers with no prices, with Conversation Builder and KnowledgeAI from Silver and the AI agents on Gold. From LivePerson's pricing page.

Generative AI tokens are an entitlement agreed with your sales representative. Proactive messages are charged per channel plus a 15% handling fee, and SMS gateway and phone number costs are extra. We found no free trial.

What buyers estimate LivePerson costs

Any number for LivePerson is an estimate, not LivePerson's price. Vendr, which collects buyers' contract data, shows no typical price for LivePerson, only a form to request one. LiveChat, a rival, says LivePerson's annual contracts often exceed $60,000, and cites no source.

If that rival's floor holds, $60,000 a year is $5,000 a month. Ask LivePerson for its real figure, with the Gold tier and your token entitlement in writing.

How Algomo bills

Algomo's free plan is self-serve, with 100 credits a month, about twenty helped conversations, and no card. After that you pick one monthly amount, from $50 for 2,000 credits up to $4,300 for 200,000, with Enterprise above. The full ladder is on [Algomo's pricing page](/pricing).

A conversation that helped someone costs 5 credits, plus 5 for every further ten replies. One that went nowhere, spam or gibberish, costs nothing. There are no seats, the amount you pick is the most you can be charged, and unused credits roll over.

Here is the sum. 4,000 helped conversations a month are 20,000 credits. The $400 plan holds 16,000, so you would buy the $600 plan, which holds 25,000.

How the two bills work

| | Algomo | LivePerson |

| --- | --- | --- |

| Prices on the website | ✓ Every plan | ✗ Tiers only, no prices |

| How you start | ✓ Free plan, no card | A quote or a request for proposal |

| AI agents included from | ✓ The free plan | Gold, the top tier |

| Generative AI usage | ✓ Priced per helped conversation | Token entitlements agreed with sales |

| A busier month than planned | ✓ Capped at the amount you set | Set by your contract |

Our take: send LivePerson your monthly conversation count and ask for a written Gold quote, then compare it with the Algomo plan that holds the same count.

How does each one protect customer data?

TL;DR: LivePerson lists SOC 2 with HIPAA, PCI DSS 4.0 and HITRUST, while Algomo's agent works from the visitor's browser and stores no token for anything your website already does.

An agent that can see orders needs a way into your systems. Your security team will ask what the vendor keeps to get that access, and what it holds.

LivePerson security

[LivePerson's security page](https://www.liveperson.com/resources/security/) lists an SSAE 18 SOC 2 report with HIPAA, PCI DSS 4.0, GDPR, SOX controls, HIPAA business associate agreements and a HITRUST audit. It describes its security management system as aligned with ISO 27001. LivePerson also lets you bring your own language model from OpenAI, Google, Meta, Cohere or Anthropic.

Algomo security

Algomo's agent runs in the visitor's browser, on your page. When it checks a price or the visitor's own order, the request leaves that browser the same way your page sends it. For all of that, Algomo stores no token or credential.

A system only your staff can reach, such as billing, needs a connector, and Algomo stores that login encrypted with AES-256-GCM. You choose which systems get one, and you can revoke it at any time.

Algomo holds Cyber Essentials certification across the whole organisation and is registered with the ICO under ZA821756. It works to UK and EU GDPR, the EU AI Act and DORA, and hosts its application, database and files in the EU. On Enterprise you choose the AI model provider, and [the data processing agreement](/dpa) covers every plan. [Algomo's compliance page](/trust) has the detail.

Connectors in Algomo for systems behind a staff login. Every stored login is encrypted and can be revoked.

Our take: ask each vendor to list what it stores that could open one of your customers' accounts.

What does the SoundHound purchase mean for a LivePerson buyer?

TL;DR: SoundHound AI completed its purchase of LivePerson on 4 September 2026 and says LivePerson's platform will be integrated into its OASYS agent system.

A platform you sign for today is one you will run for years. When the vendor changes hands, your contract, roadmap and account team are worth confirming in writing.

SoundHound and LivePerson signed the deal in April 2026 and amended it in July, together with an agreement to swap LivePerson's secured notes for SoundHound shares and cash. Shareholders approved it on 2 September, and SoundHound announced completion two days later. [SoundHound's announcement](https://www.sec.gov/Archives/edgar/data/0001840856/000121390026097712/ea030469101ex99-1.htm) says LivePerson's debt was retired, its shares left the Nasdaq, and its platform will be integrated into OASYS. LivePerson's website still sells Conversational Cloud under the LivePerson name.

Our take: ask your LivePerson account team which products and names your contract covers after the move to OASYS, and get the answer in writing.

What to ask any vendor selling you an AI agent

TL;DR: Five questions, each with a test you can run during the demo.

- Who connects the agent to our order system, and how many days does it take? Ask for a dated plan. - Does it read our live website, or an imported copy? Ask where one answer came from. - If we change a price now, what does it say in ten minutes? Change one and ask. - Can it do the step for the visitor, such as adding an item to the basket? Ask it to. - Who carries the risk of a wrong answer? Ask to see it in the contract.

Run every test on your own website, never on the vendor's demo.

Which one should your team choose?

Choose LivePerson if

- Most of your conversations arrive on WhatsApp, SMS, Apple Messages or the phone.

- You run a contact centre and want the AI, the human agents and the analytics on one platform.

- Your procurement needs a SOC 2 report, PCI DSS or a HITRUST audit from the vendor today.

Don't choose LivePerson if

- Your questions start on your website, and the answers live on its pages and the systems behind them.

- You have no developer free to define each API call, or no budget for a services team.

- Your answers must match today's website, not the last sync of a knowledge base.

- You want to see the price before a sales cycle, and AI agents on every plan.

- You want the agent to do the step on the page with your visitor.

Conclusion

TL;DR: Choose LivePerson when you need one contact-centre platform across messaging apps and voice, and choose Algomo when your customers are on your website and you want the integration done for you and every answer read live.

On the cost of the integration, LivePerson gives your developers the tools to define each call, with services on offer. Algomo finds the systems behind your site and connects to them itself, with a first version the same day.

On whether answers stay true, LivePerson answers from a knowledge base you import and sync, and its docs say you accept the hallucination risk of AI-written answers in a customer-facing bot. Algomo reads the live page and backs a reviewed agent with a free month.

Questions buyers ask about Algomo and LivePerson

Who owns LivePerson?

SoundHound AI owns LivePerson. The merger closed on 4 September 2026, and [LivePerson's merger filing](https://www.sec.gov/Archives/edgar/data/1102993/000119312526383671/d112155d8k.htm) calls it an indirect wholly owned subsidiary of SoundHound. Its shares left the Nasdaq and its debt was retired. SoundHound says LivePerson's platform will be integrated into OASYS, its agent system.

How much does LivePerson cost?

LivePerson publishes no prices. Its pricing page lists Bronze, Silver and Gold tiers and asks for a quote or a request for proposal, with the AI agents on Gold. Vendr shows no typical figure, and LiveChat, a rival, says contracts often exceed $60,000 a year. That figure is an estimate, not LivePerson's price.

Is Algomo a good LivePerson alternative?

Yes, when your customers' questions start on your website. Algomo connects itself to the systems behind your pages, answers from the live page, and shows every price on its pricing page. If most of your volume arrives on messaging apps or by phone, LivePerson covers those channels and Algomo does not.

Can LivePerson's AI read my website?

LivePerson's current import article lists CSV files, Google Sheets and a synced content system as the ways to fill its knowledge base. It lists no web page import. Algomo opens your live page for every question.

Does LivePerson have a free trial?

We found none. The pricing page offers a quote or a request for proposal. Algomo's free plan has no card and no end date, with 100 credits a month.

Can I move from LivePerson to Algomo?

Yes, for the website channel. Give Algomo your site address and it finds what your pages call, with a first version the same day. Past tickets can be saved into Algomo's knowledge base, and a connector can create tickets in your helpdesk when a person needs to take over.

Which AI model does Algomo use?

The default model is from OpenAI. On Enterprise you choose the provider, and the application, database and files stay hosted in the EU.

Where these facts come from

The sources are:

- [LivePerson's home page](https://www.liveperson.com/), [LivePerson's AI agents page](https://www.liveperson.com/products/ai-agents/) and [LivePerson's AI agent builder page](https://www.liveperson.com/products/ai-agent-builder/): the product, the agent types and AI Studio. - [LivePerson's integrations page](https://www.liveperson.com/products/integrations/) and [LivePerson's professional services page](https://www.liveperson.com/services/professional-services/): connectors, Workato and services. - [The API integrations article](https://developers.liveperson.com/conversation-builder-integrations-api-integrations.html): how a bot calls your systems, and the timeout. - The [import content article](https://developers.liveperson.com/knowledgeai-internal-knowledge-bases-import-content.html), the [KMS article](https://developers.liveperson.com/knowledgeai-internal-knowledge-bases-integrate-a-kms-or-cms.html) and the [KnowledgeAI agents article](https://developers.liveperson.com/conversation-builder-generative-ai-knowledgeai-agents-automate-enriched-answers.html): where answers come from, the sync and the hallucination risk. - The [escalation article](https://developers.liveperson.com/conversation-builder-integrations-liveperson-agent-escalation-integrations.html) and the [Preview article](https://developers.liveperson.com/conversation-builder-testing-deployment-preview-the-bot.html): handoff and testing. - [LivePerson's Syntrix page](https://www.liveperson.com/syntrix/): simulation and LivePerson's own testing estimate. - [LivePerson's messaging channels page](https://www.liveperson.com/products/messaging-channels/) and [LivePerson's voice page](https://www.liveperson.com/products/voice-ai/): channels. - [LivePerson's pricing page](https://www.liveperson.com/pricing/): tiers, AI agents on Gold, tokens and fees. - [Vendr's LivePerson page](https://www.vendr.com/marketplace/liveperson), and [LiveChat's LivePerson page](https://www.livechat.com/comparison/liveperson-alternative/), a rival's estimate. - [LivePerson's security page](https://www.liveperson.com/resources/security/): the certificate list. - [SoundHound's announcement](https://www.sec.gov/Archives/edgar/data/0001840856/000121390026097712/ea030469101ex99-1.htm), [LivePerson's merger filing](https://www.sec.gov/Archives/edgar/data/1102993/000119312526383671/d112155d8k.htm) and [LivePerson's shareholder approval release](https://www.sec.gov/Archives/edgar/data/0001102993/000119312526380467/d14310dex991.htm): the purchase. - [Algomo's pricing page](/pricing), [Algomo's compliance page](/trust), [the data processing agreement](/dpa) and [the guarantee page](/resources/guarantee). - [The demo library](/demos): [the Blinds 2go demo](/demos/blinds2go), [the Nike demo](/demos/nike) and [the Qualtrics demo](/demos/qualtrics).

All the information on this page was available in November 2026.

Served to agents at /compare/vs-liveperson.md, or at this URL with an Accept: text/markdown header.