15 Best Conversational AI Platforms for 2026

TL;DR Summary 

Conversational AI platforms enable businesses to automate multi-step customer interactions across voice, chat, and messaging channels far beyond rule-based chatbot capabilities.

• Gartner projects conversational AI will reduce contact center labor costs by $80 billion in 2026, automating one in ten agent interactions.

• The global conversational AI market reached approximately $17.12 billion in 2026, growing at 21–25% annually according to Fortune Business Insights.

• Kore.ai leads enterprise deployments because well-tuned implementations achieve containment rates in the 60–70% range, reducing live-agent volume significantly.

• Intercom Fin and Tidio Lyro are GenAI-native platforms, resulting in stronger multi-step handling and context-aware handoffs than rule-based legacy tools.

• Intercom, Zendesk, and Gorgias have shifted to per-resolution pricing, aligning vendor incentives with customer outcomes rather than seat counts.

Introduction

Imagine it is 11 PM on the Friday of Black Friday weekend. A shopper has three items in their cart and has one question about whether the order will arrive before a birthday. Your support queue is closed. Your bot returns a link to a shipping policy page. The shopper closes the tab. That sale is gone, and your team will not see it in a report until Monday.

This is the gap a modern conversational AI platform is built to close. The category has moved well past curiosity. Gartner projected that conversational AI deployments would reduce contact center agent labor costs by $80 billion in 2026, driven by one in ten agent interactions being automated. For ecommerce brands in particular, the shift from manual queues to AI-assisted support has moved from an experiment to a budget line. This guide separates platforms by who they are actually built for and scores each one on a 5-point Agentic AI scale.

What Is a Conversational AI Platform, and How Is It Different From a Chatbot?

A conversational AI platform is a SaaS product that lets businesses build, deploy, and manage AI-powered agents across text, voice, and messaging channels. It goes beyond scripted chatbot trees to understand intent, handle multi-step tasks on its own, and hand off to humans with full context. According to Fortune Business Insights, the conversational AI market is valued at approximately $17.12 billion in 2026 and is growing at around 21-25% annually.

The distinction between a chatbot and conversational AI platform matters more in 2026 than it did two years ago. A chatbot is a rule-based decision tree: it matches keywords to a finite set of prewritten paths, and the moment a customer phrases something differently, it breaks down. A conversational AI platform uses natural language understanding and large language models to read intent, hold context across a session, and act. If you are comparing platforms on response speed alone, you are measuring the wrong variable.

Gartner's working definition covers SaaS products that enable multi-channel simulation of human conversation using a combination of composite AI and generative AI. In simple terms, a chatbot answers a question and stops there. A conversational AI platform goes further: it understands the context behind the question, takes the necessary steps to resolve the issue, and hands it off to a human agent with full context when it cannot. That difference is what determines whether a customer leaves the conversation satisfied or opens another ticket.

How Conversational AI Platforms Are Changing in 2026

The conversational AI market has shifted meaningfully in the past two years, and the platforms available today behave very differently from those that were standard in 2023. These changes are shaping which platforms are worth evaluating and which are not.

GenAI-native versus GenAI-layered architecture

Not all platforms are built the same way under the hood. Some, including Intercom Fin and Tidio Lyro, were built on large language models from the start. Others added generative features on top of an existing rule-based engine without rebuilding the core. In normal conditions, the two can look similar. The difference tends to surface when a customer asks something unexpected, when a conversation spans multiple steps, or when a handoff to a human agent needs to carry full context. A platform built on a rule-based foundation may struggle in those moments in ways that are not obvious from a demo.

The move toward agentic task completion

The expectation for what these platforms should do has shifted. A few years ago, handling a high share of FAQ queries on its own was considered a strong result. Today, buyers expect platforms to go further: modifying an order, initiating a return, recovering an abandoned cart, or sending a proactive shipping alert without a human step. Platforms that are still optimized only for answering questions are losing ground, particularly in ecommerce, where a single customer conversation can span chat, a delivery update, and a return request within the same session.

A note on pricing models

It is also worth understanding how pricing has changed before entering any vendor conversation. Intercom, Zendesk, and Gorgias have all moved toward per-resolution pricing over the past year, moving away from per-seat licensing. Under this model, you pay for issues that are actually resolved rather than for the number of agents using the platform. That shift aligns the vendor's incentive more closely with yours and is worth factoring in when comparing costs across shortlisted platforms.

Platform Comparison: All 15 at a Glance

How We Evaluated These 15 Platforms

Here is exactly how we scored each platform so that you can weigh the conclusions yourself. Seven dimensions were assessed:

• NLU accuracy and depth of intent recognition

• Agentic AI capability, scored 1-5 on the scale defined later in this guide

• Omnichannel coverage, including WhatsApp, Instagram DM, voice, and live chat

• Native ecommerce integrations (Shopify, WooCommerce, Magento, Gorgias), not Zapier bridges

• Deployment timeline: time-to-first-bot-live and time-to-full-production

• Pricing transparency and total cost of ownership across three volume tiers

• Human escalation quality: warm transfer, context-passing, and cold-transfer handling

Sources include vendor documentation, public case studies, G2 and Gartner Peer Insights review analysis, and published benchmark data, where available.

Note: No platform paid for placement. Where a vendor ranks itself first in its own guide, we flag it rather than inherit the ranking.

The 15 Best Conversational AI Platforms for 2026

The platforms below cover a range of ecommerce and enterprise support needs, from social commerce automation and WISMO resolution to omnichannel contact center orchestration. Each review is structured the same way so you can compare them on the same basis. Pricing and feature details change frequently, so treat every figure as a directional starting point and confirm current numbers with the vendor before committing.

1. Kore.ai

Best for: Enterprise CX and contact centers

Agentic AI: 5/5

Starting price: Contact for pricing

Channels: Voice, chat, WhatsApp, email, IVR

Kore.ai is the platform most often cited by buyers when discussing genuine enterprise-grade agentic automation. It pairs deep NLU with orchestration across systems, which puts it in the higher band of tier-1 containment, commonly cited in the 60-70% range for well-tuned deployments.

Strengths: Strong multi-step automation; mature voice and digital parity; flexible LLM and SLM orchestration.

Limitations: Steeper learning curve; meaningful configuration effort; pricing requires a sales conversation.

Ecommerce fit: Capable but heavier than most DTC teams need; better suited to large retailers with internal ops resources.

Time to deploy: Weeks to a few months for full production.

2. Intercom Fin

Best for: Fast deployment, no-code setup

Agentic AI: 4/5

Starting price: Outcome-based (per resolution)

Channels: Web, in-app, email, messaging

Fin is one of the clearest examples of a GenAI-native agent with outcome-based pricing: you pay for resolutions, not seats. Many teams live in hours rather than weeks, making it a natural first entry in the shortlist for lean support teams.

Strengths: Fast time-to-live; pricing aligned to resolved outcomes; strong out-of-the-box accuracy.

Limitations: Per-resolution cost can climb at volume; deepest value assumes the full Intercom suite; voice is not a strength.

Ecommerce fit: Good for support-heavy DTC, though native commerce actions are lighter than Gorgias.

Time to deploy: Hours to days.

3. Yellow.ai

Best for: Enterprise omnichannel with commerce connectors

Agentic AI: 4/5

Starting price: Contact for pricing

Channels: Voice, chat, WhatsApp, social, email

Yellow.ai sits between enterprise scale and commercial practicality, with a broad channel footprint and connectors that integrate with retail stacks. It suits mid-to-large brands that want one platform for support and marketing.

Strengths: Wide channel coverage; strong in APAC and emerging markets; commerce-aware flows.

Limitations: Configuration depth can be demanding; quality varies by use case; pricing is not transparent.

Ecommerce fit: Solid Shopify and marketplace connectors for larger retailers.

Time to deploy: Days to weeks.

4. Cognigy

Best for: Omnichannel contact centers

Agentic AI: 5/5

Starting price: Contact for pricing

Channels: 100+ connectors, including voice, chat, and WhatsApp

Cognigy's strength is its breadth: 100+ channel connectors and genuine voice-digital parity, which is why it often appears on enterprise contact-center shortlists. Its orchestration handles complex flows without losing context on handoff.

Strengths: Exceptional channel coverage; strong voice automation; enterprise governance tools.

Limitations: Built for scale, so more than most SMBs will use; requires technical owners; sales-led pricing.

Ecommerce fit: Suited to large retailers; more than most DTC brands will need.

Time to deploy: Weeks.

5. Tidio Lyro

Best for: DTC ecommerce

Agentic AI: 4/5

Starting price: Free tier; paid from approximately $29/month

Channels: Web chat, WhatsApp, Instagram, email

Lyro is Tidio's AI agent that’s built for online retail. Having native Shopify integration, WhatsApp commerce, and built-in cart recovery make it the default starting point for growing DTC brands.

Strengths: Easy setup; affordable entry; strong ecommerce templates.

Limitations: Less suited to complex enterprise flows; advanced agentic actions require higher tiers.

Ecommerce fit: Excellent. Native Shopify, WhatsApp, and cart recovery out of the box.

Time to deploy: Hours.

6. Gorgias AI

Best for: Shopify-first DTC support

Agentic AI: 4/5

Starting price: From approximately $10/month base; AI add-ons vary

Channels: Web chat, email, WhatsApp, social

Gorgias is built around ecommerce with native Shopify, WooCommerce, and Magento connections, and an AI agent that can take order actions rather than just answer questions. For DTC teams already using a helpdesk, it is the most natural upgrade path.

Strengths: Deep native ecommerce integrations; order-aware automation; post-purchase use cases.

Limitations: Costs add up with ticket volume; best value inside the Gorgias ecosystem; voice is limited.

Ecommerce fit: Among the strongest. Built for Shopify-centric retailers.

Time to deploy: Hours to days.

7. ManyChat

Best for: Social commerce and messaging

Agentic AI: 3/5

Starting price: Free tier; paid from approximately $15/month

Channels: Instagram DM, WhatsApp, Messenger, SMS

ManyChat is the most widely used platform for social-commerce conversations, particularly on Instagram and WhatsApp. It is less an agentic problem-solver and more a high-converting messaging automation layer, which is what many DTC marketing teams actually want.

Strengths: Best-in-class social and messaging flows; strong marketing automation; low entry cost.

Limitations: Lighter on true support resolution; limited deep system actions; not a full helpdesk.

Ecommerce fit: Excellent for social-led DTC; pair with a support tool for full resolution.

Time to deploy: Hours.

8. Freshdesk Freddy

Best for: SMB budgets and Freshdesk users

Agentic AI: 3/5

Starting price: Free tier available

Channels: Web chat, email, WhatsApp, voice add-ons

Freddy is Freshworks' AI layer, and its appeal lies in its reach alongside a genuine free tier. For teams already on Freshdesk, escalation is native, and the ramp is straightforward, making it a sensible no-code starting point for cost-conscious teams.

Strengths: Free entry point; native Freshdesk escalation; broad feature set.

Limitations: Best value inside the Freshworks ecosystem; agentic depth trails the leading platforms; commerce actions are limited.

Ecommerce fit: Workable through integrations; not e-commerce-native.

Time to deploy: Hours to days.

9. Zendesk AI

Best for: Existing Zendesk customers

Agentic AI: 4/5

Starting price: Add-on to Zendesk Suite

Channels: Web, email, messaging, voice

Zendesk has moved toward resolution-based AI, and for brands already on the suite, the upgrade is straightforward. Its agents handle multi-step intents well and pass a clean context on handoff, which is a real advantage at scale.

Strengths: Mature platform; strong context-passing on handoff; resolution-based pricing options.

Limitations: Costs scale quickly; full value assumes the Zendesk suite; requires configuration depth.

Ecommerce fit: Good through apps; not Shopify-native out of the box.

Time to deploy: Days to weeks.

10. LivePerson

Best for: Large-scale messaging operations

Agentic AI: 4/5

Starting price: Contact for pricing

Channels: Messaging, voice, web, social

LivePerson is a veteran in conversational commerce and messaging at scale. It suits enterprises that treat messaging as a primary revenue channel rather than a support afterthought.

Strengths: Deep messaging expertise; detailed analytics; enterprise reliability.

Limitations: Enterprise pricing and complexity; longer onboarding; heavier than small DTC needs.

Ecommerce fit: Strong for large retailers; more than most small DTC brands will use.

Time to deploy: Weeks.

11. IBM Watsonx Orchestrate

Best for: Regulated enterprises and internal workflows

Agentic AI: 4/5

Starting price: Contact for pricing

Channels: Chat, voice, internal apps

IBM's WatsonX Orchestrate leans into agentic automation across customer and internal workflows, with governance and security that appeal to regulated industries. It fits where compliance comes before speed-to-launch.

Strengths: Strong governance and security; broad enterprise integrations; serious agentic ambitions.

Limitations: Complex to implement; enterprise-only economics; not aimed at DTC.

Ecommerce fit: Limited DTC relevance; better for large internal-plus-external use cases.

Time to deploy: Weeks to months.

12. Google Dialogflow CX

Best for: Developers building custom agents

Agentic AI: 3/5

Starting price: Usage-based (per request)

Channels: Voice, chat, custom via API

Dialogflow CX is a builder's platform: flexible, usage-priced, and deeply integrated with Google Cloud. It rewards teams with engineering resources and does not suit those expecting an out-of-the-box agent. Agentic capability depends largely on what you build on top of it.

Strengths: Highly customizable; pay-as-you-go pricing; strong Google Cloud integration.

Limitations: Requires development work; little turnkey ecommerce value; ongoing maintenance overhead.

Ecommerce fit: Only through a custom build; not a packaged DTC solution.

Time to deploy: Weeks, development dependent.

13. Rasa

Best for: Open-source, full-control deployments

Agentic AI: 3/5

Starting price: Open source; Pro and enterprise tiers available

Channels: Custom via framework

Rasa is the leading open-source framework for teams that want full control over their data and behavior. It works well for governance and on-premises control. Note: Rasa places itself first in its own published guide; our independent assessment places it in the top tier for open-source governance, not overall packaged CX capability.

Strengths: Open source and self-hostable; maximum data control; strong privacy posture.

Limitations: Requires ML and development expertise; slow time-to-value; no out-of-the-box commerce functionality.

Ecommerce fit: Custom builds only; not suited to non-technical DTC teams.

Time to deploy: Weeks to months.

14. Drift (Salesloft)

Best for: B2B conversational marketing and sales

Agentic AI: 3/5

Starting price: Contact for pricing

Channels: Web chat, email, conversational marketing

Now part of Salesloft, Drift is built for pipeline: qualifying visitors, booking meetings, and routing leads in real time. It is a B2B revenue layer rather than a DTC support tool.

Strengths: Excellent for lead capture and routing; strong sales-stack integration; conversion focus.

Limitations: Limited support resolution depth; B2B-oriented; not ecommerce-native.

Ecommerce fit: Weak for DTC support; better for B2B and considered-purchase sites.

Time to deploy: Days to weeks.

15. Sendbird

Best for: In-app messaging and AI agents at scale

Agentic AI: 4/5

Starting price: Usage-based; contact for enterprise

Channels: In-app chat, messaging, AI agents

Sendbird is strongest where conversations live inside your own app. Its developer-friendly APIs suit product-led brands that want embedded messaging rather than a bolt-on widget.

Strengths: Excellent in-app messaging; developer-friendly; scalable AI agents.

Limitations: Requires development to realize full value; less suited to standard storefronts; pricing scales with usage.

Ecommerce fit: Good for app-first commerce; less so for standard Shopify storefronts.

Time to deploy: Days to weeks, depending on development.

Looking across all 15 platforms, a few patterns emerge. Enterprise buyers with complex contact center requirements tend to gravitate toward Kore.ai and Cognigy, both of which score 5/5 on the agentic scale and are built for multi-system orchestration at scale. DTC and ecommerce brands with Shopify-centric stacks will find the strongest native fit in Tidio Lyro and Gorgias AI. And teams that need to go live quickly without engineering support consistently shortlist Intercom Fin for its deployment speed and outcome-based pricing. Everything else in the list serves a more specific buyer profile, which the individual reviews cover in detail.

Agentic AI Scorecard: What Each Score Means in Practice

The 5-point scale below was used throughout this guide. Match the score to the job you actually need done.

Only platforms scoring 4 or 5 can recover an abandoned cart at 2 AM or send a proactive delivery notification without a human in the loop. A platform that scores 2 cannot, regardless of how the demo looks.

Platforms Built for Ecommerce and DTC Brands

If you run a DTC brand, operate a Shopify store, or lead ecommerce CX, this section covers the questions that decide whether a conversational AI platform actually fits online retail.

1. Native commerce integration, not Zapier bridges

Tidio Lyro and Gorgias AI read order data directly from Shopify, WooCommerce, and Magento. Some platforms connect to these systems through a third-party automation tool like Zapier rather than a direct integration. That indirect connection introduces latency, which is a problem when a customer is asking about their order and expects an answer in seconds. Handling WISMO queries is the highest-volume use case for ecommerce support, and a direct data connection is what makes it work reliably.

2. WhatsApp, Instagram DM, and TikTok Shop as primary channels

For DTC brands, a significant share of customer conversations happen on social and messaging channels rather than a website chat widget. Tidio and Gorgias cover the major messaging channels as well. Where it gets relevant for buyers is how deeply these channels are integrated.

Platforms that were built for large enterprise contact centers first and expanded into social messaging later often treat WhatsApp and Instagram as add-ons, which means the flows, data access, and escalation logic may not be as seamless as they appear in a demo. For a DTC brand where a meaningful share of orders and support queries come through Instagram or WhatsApp, that distinction is worth testing before you commit.

3. Seasonal autoscaling

Black Friday and Cyber Monday can bring ten times the normal conversation volume within hours. A well-built conversational AI platform absorbs that automatically, without any drop in response time or resolution quality. Not all platforms handle this equally well. Smaller or less mature platforms running on shared infrastructure can slow down or fail to trigger escalations correctly under heavy load. These limitations tend to be invisible during normal traffic and only surface during peak periods. When evaluating vendors, ask specifically how their infrastructure handles sudden volume spikes and whether SLA guarantees apply during peak periods. A vendor that cannot answer clearly is worth approaching with caution.

4. Post-purchase use cases

The strongest ecommerce brands use conversational AI beyond reactive support. Proactive shipping updates, return initiation, and post-delivery follow-up are where the real retention value sits. A self-serve returns flow triggered inside a chat conversation lets the customer resolve the issue without opening a ticket, which reduces support volume and leaves the customer with a better last impression of the brand. That combination, fewer tickets and a smoother experience, is what makes post-purchase automation worth building.

5. Abandoned cart recovery

Cart abandonment is one of the more measurable use cases for conversational AI in ecommerce. DTC brands using chat and WhatsApp recovery flows alongside email report higher conversion rates than those relying on email alone, largely because the intervention happens in real time while the shopper is still considering the purchase. That said, results vary significantly depending on trigger timing, cart value threshold, and message quality. Before relying on vendor benchmarks, run the flow against your own traffic and build a baseline from your actual numbers.

KPIs to Track for Ecommerce Conversational AI

• Containment rate: the share of conversations AI resolves without any human involvement

• Cart recovery rate: the percentage of abandoned carts converted through a conversational prompt

• Post-purchase CSAT: satisfaction measured specifically in the window between order placement and delivery

• Average resolution time: how long it takes from the first message to a closed conversation

• Revenue influenced per conversation: the direct and assisted revenue attributable to AI-handled interactions

If a vendor cannot report on these metrics, you cannot build a reliable ROI case.

Pricing and Total Cost of Ownership

'How much does a conversational AI platform cost?' rarely has a single answer because the visible license is often half the actual cost. The table below gives a total-cost view across three volume tiers.

Three pricing models dominate the market. Per-seat licensing is the legacy model and is being phased out. Per-conversation or per-resolution is the current standard, used by Intercom Fin and Gorgias AI. Outcome-based pricing is the emerging model: you pay only for verifiably resolved issues, which aligns vendor incentives with yours.

Budget separately for the costs that vendor pages tend not to surface: per-conversation overages, telephony fees on voice, LLM API call costs, integration fees, training-data preparation, and model fine-tuning. Together, these can roughly double the headline platform price. Confirm current figures with each vendor before committing.

How to Choose a Conversational AI Platform: A 6-Step Framework

The most common selection mistake is evaluating features before defining the primary channel. Run these six steps in order.

Step 1: Define your primary channel first

WhatsApp and Instagram point to DTC ecommerce tools. Voice points to contact-center platforms. Web chat points to CX and helpdesk tools. True omnichannel points to enterprise-grade platforms. Voice-first and chat-first architectures differ at every level, so this decision shapes every subsequent cost and integration choice.

Step 2: Score platforms on agentic capability for your use case

Use the 1-5 scale above. FAQ containment needs only a 2. Cart recovery and proactive outreach need a 4 or higher. Buying a 5 to answer FAQs wastes budget; buying a 2 to recover carts wastes the opportunity.

Step 3: Map your tech stack before evaluating integrations

List your CRM, helpdesk, ecommerce platform, and data warehouse. Check for native connectors (Shopify, Gorgias, Salesforce, Zendesk) rather than Zapier workarounds, because real-time data access depends on how deep the connector goes, not whether a logo appears on a partner's page.

Step 4: Pressure-test deployment timelines

Ask for case studies showing time-to-first-bot-live and time-to-full-production, not sales estimates. Intercom Fin can go live in hours; Kore.ai typically requires weeks of configuration and training data work. Both are reasonable, but price the timeline honestly.

Step 5: Model the total cost of ownership across three years, not just the monthly license

Include implementation, training-data cost, integration development, overage risk, and annual model-update fees. A lower-priced platform with heavy overages often costs more on a three-year view.

Step 6: Test human escalation quality, not just bot capability

Warm transfer, cold transfer, and context-passing protocols determine the experience at the exact moment the bot cannot resolve an issue. Ask vendors directly: what context passes to the human agent, and in what format? The handoff is what separates a good customer experience from a frustrating dead end.

Conclusion

There is no single best conversational AI platform, only the best one for your channel, your stack, and the tasks you need automated. Ecommerce and DTC brands should start with Tidio Lyro, Gorgias AI, and ManyChat. Enterprise CX teams should look at Kore.ai, Cognigy, and Intercom Fin. No-code buyers should shortlist Intercom Fin and Freshdesk Freddy.

Whatever you choose, score it on agentic capability first, model the three-year total cost honestly, and test the human handoff before signing. The handoff is where most implementations succeed or fail.

For ecommerce brands, a large share of conversations are about orders, shipping, and returns. The conversational layer and the post-purchase data layer need to work together. If post-purchase is where your tickets pile up, it is worth understanding how a connected shipping intelligence platform feeds your conversational AI with live tracking, estimated delivery dates, and exception data.

Frequently Asked Questions

What is the best conversational AI platform for Shopify?

Tidio, Lyro, and Gorgias AI are the three strongest options for Shopify-based brands. Both connect natively to Shopify order data, support WhatsApp and Instagram, and handle WISMO queries without a Zapier bridge. Tidio is the better entry point in terms of price and setup speed; Gorgias is better for teams already using it as their helpdesk.

How much does a conversational AI platform cost?

Pricing varies significantly by volume and feature tier. SMB plans start at free or low double digits per month; enterprise contracts are typically negotiated and can run into thousands per month. Request pricing directly from any vendor before budgeting.

What is the difference between a chatbot and a conversational AI platform?

A chatbot follows fixed decision-tree scripts and handles only scenarios it has been explicitly programmed for. A conversational AI platform reads free-text intent through natural language processing, holds context across a session, and in agentic implementations, takes multi-step actions without a human step. Most mature setups combine both: rules for predictable tasks, and AI for everything less structured.

What is the best no-code conversational AI platform for a small business?

Intercom Fin and Tidio Lyro are the strongest no-code options. Both can go live in hours, offer free or low-cost entry points, and do not require engineering to configure. Freshdesk Freddy is a good alternative for teams already on Freshdesk.

 Can a conversational AI platform automatically recover abandoned carts?

Yes, platforms scoring 4 or 5 on the agentic scale can trigger cart-recovery conversations via on-site chat, WhatsApp, or SMS based on exit-intent signals or inactivity thresholds. Tidio Lyro, Gorgias AI, and ManyChat all support this. Results vary significantly by trigger timing, AOV threshold, and message quality, so test against your own baseline before relying on vendor benchmarks.

What is agentic AI in customer service?

Agentic AI refers to systems that complete multi-step tasks autonomously rather than just answering questions. In customer service, this means a system that can detect a delivery exception, proactively notify the customer, rebook the shipment, issue a credit, and close the ticket without any human involvement. Platforms scoring 5/5 on the agentic scale can do this; those scoring 2 or 3 cannot.

How long does it take to deploy a conversational AI platform?

It depends on the platform and what you are building. Intercom Fin and Tidio Lyro can go live in hours. Gorgias AI and ManyChat take hours to days to complete. Zendesk AI and Yellow.ai take days to weeks to deliver results. Kore.ai, Cognigy, and IBM WatsonX take weeks to months for full production deployments.