Amazon Comprehend is a natural language processing (NLP) service that uses machine learning to find insights and relationships in text. Amazon Comprehend uses machine learning to help uncover insights and relationships in unstructured data. The service identifies the language of the text; extracts key phrases, places, people, brands, or events; understands how positive or negative the text is; analyzes text using tokenization and parts of speech; and automatically organizes a collection of text…
$0
per unit
Amazon Lex
Score 7.5 out of 10
N/A
AWS provides Amazon Lex, a chatbot building technology.
$0
Per Speech Request
Fin by Intercom
Score 8.7 out of 10
N/A
Fin is Intercom’s AI Agent for customer service, designed to deliver high-quality answers, even for complex queries. It works with any helpdesk, or it can be paired with Intercom’s next-generation Helpdesk to get the full Intercom Customer Service Suite.
$0.99
one-time fee per outcome
Pricing
Amazon Comprehend
Amazon Lex
Fin by Intercom
Editions & Modules
Syntax Analysis
$0.00005
per unit
Key Phrase Extraction
$0.0001
per unit
Sentiment Analysis
$0.0001
per unit
Entity Recognition
$0.0001
per unit
Language Detection
$0.0001
per unit
Pll Detection
$0.0001
per unit
Event Detection Per Event Type
$0.003
per unit
Request and Response
$0.004
Per Speech Request
Stream Conversation
$0.0065
Per Speech Interval
Fin with your current helpdesk
$0.99
one-time fee per outcome
Copilot add-on
$35
per month per user
Pro
$99
per month For analysis of 1,000 conversations
Fin with Intercom’s Helpdesk
from $39 + $0.99 per Fin outcome
per month per seat
Offerings
Pricing Offerings
Amazon Comprehend
Amazon Lex
Fin by Intercom
Free Trial
Yes
No
Yes
Free/Freemium Version
No
No
No
Premium Consulting/Integration Services
No
No
No
Entry-level Setup Fee
No setup fee
No setup fee
No setup fee
Additional Details
—
—
Fin comes with a 90-day money-back guarantee. Here's how it works:
Intercom states that users who sign up for the Fin Guarantee Success Program and do not achieve at least a resolution rate of 65% will be paid $1M. This program is designed for high volume customers.
Eligibility criteria:
High volume customers (over 250k monthly conversions) in North America and Europe. Intercom states that phase one of this program will admit customers on Intercom Helpdesk or Zendesk.
Specifically, it starts processing millions of documents in minutes by leveraging the power of machine learning without having trained models from scratch. If any of the content contains personally identifiable information not only can Amazon Comprehend locate it but it will also redact or mask it. Using NLP techniques Amazon Comprehend goes well beyond keyword search or rules-based tagging to accurately classify documents. For my task or development, I cannot find any difficulties with Amazon Comprehend.
If you wish to quickly deploy multilingual chatbots without having to worry about infrastructure and model training, go for Amazon Lex. It is one of the best general-purpose conversational AI solutions in the market. The cherry on the cake is that it also seamlessly integrates with other AWS services, so you would be good there. Performance monitoring is very easy with AWS. It has support for both text and integration. If you are not a pro-NLP expert, Amazon Lex will make your job really easy.
Fin is great for using for first line support. We use Fin for conversations where customers have a standard question, and Fin is able to pull from our content to answer this accurately and go above and beyond to include some basic problem solving. This really helps us free up our Support Team's time for more complex queries. Fin isn't appropriate for us on technical issues or conversations which require human support. We've had to remove Fin from interacting on these conversations as customers were becoming frustrated with speaking to AI, or having Fin be unable to problem solve. However, this was easy to set up through Intercom, and now customers with complex questions/situations or bugs/technical issues do not engage with Fin and Fin only handles suitable conversations.
Amazon Comprehend identifies the language of the text and extracts Key-phrases, places, people, brands or events.
It can build a custom set of entities or text classification models that are tailored uniquely to the organisation's need
Amazon Comprehend's medical can be used to identify medical conditions, medications, dosages, strength and frequencies from sources like doctor's notes, clinical trial reports and patient health records. This service is very good and with well an accuracy or confidence score.
It seems some users really struggle to figure out how to escalate to a human (especially through email).
Not excited about how "soft" resolutions still count as resolutions and are paid for. Though some abandoned cases appear to be able to be concluded as "the user got the answer they needed", there are others where they clearly didn't, because they just open up another chat (or even more), trying to get more info. This pads the resolution stats and makes it seem more effective than it actually is.
Cost -- Fin is quite expensive. It helps us with scaling coverage, but we're not really saving money.
We have been and will be continuing our journey with Intercom and nothing too concerning has happened that I have experienced or heard of that has us on the edge yet. If it ever happens it will be something along the lines of "Outgrowing" the use of need of the platform.
Easy to deploy and very easy to integrate with other AWS services. Automating simple tasks is also very easy with Amazon Lex. We never had NLP experts in our team, but we were still able to deploy chatbots for our support functions with minimal issues. Native integration with other AWS services like S3 and Lambda has been of paramount importance.
From an administration standpoint, Fin is very easy to set up, train, and test. Having the ability to impersonate a user in our system to see how Fin responds is huge. It lets you test several situations and throw curveballs at it (as customers will) without the risk of setting Fin live and wondering what will happen. It's also easy to fine-tune. Some chatbots you can never quite get right without spending hours on, but Fin usually takes a few minutes to dial it in. From a customer standpoint, Fin couldn't be easier to engage with. We tell customers up front it's an AI bot and they're wow'd with the experience
Community support for Amazon Lex is good. Also, since it is an AWS service, the support has a similar standard as other AWS services. We have had a couple of instances of our bots weren't able to interact with our web apps. We reached out to the support team, and they were able to resolve our issue in no time. The documentation from the Amazon Lex team also makes creating chatbots a breeze.
Intercom is the premier customer support/engagement model and it definitely has one of the top tier customer support teams as well. I don't think I have ever waited more than 5 minutes to get the information I need or get help with an issue. They are incredible and I aim to model our customer service department after them.
For natural language processing tasks or techniques, there are many service providers out there in the market such as Azure Cloud Services, IBM Watson and Google Cloud Platform (GCP), but compared with them, Amazon Comprehend is the best service provider in contents of accuracy, speed of processing multilingual text, supporting SDK for most of the languages and well documented.
We only used the free live chat version from HubSpot, so Intercom is yards better! If I were just comparing an actual live chat between the two tools, HubSpot was often clunky and delayed, and it was hard to find past conversation information
It supports better and accurately as compared with our existing or old implementations. So, we fulfil our needs as per clients' requirements and it will help to grow or improve client satisfaction.
For these specific requirements, we do not require any machine learning engineers or related professionals to hire in our organisation.
None of any negative sides can be affected our business or distract existing clients.
New role opportunities — Using the “Fin-first” approach has reduced the workload for our Tier 1 team, giving them more time to focus on their own career growth. It’s also opened the door to a dedicated, AI-focused role, where a team member regularly reviews Fin’s answers and makes updates to help it perform even better.
Enabling Fin has also reduced our response time and allowed us to meet SLA's.
