Lingo Munger is a cloud-based REST API, which parses words(a text string) and returns its meaning (a json object or xml).
We work with you to identify the user conversations, the words, and meanings, then code them up into custom rules.
Then we provide a fully hosted, secure, lightening fast API service to handle the requests and responses.
Typically, your servers receive web, voice, chat and app traffic, containing sentences of natural language text to be analysed.
Your servers send this text to Lingo Munger via secure HTTPS in real-time, which returns a json object or xml for your servers to action (eg. book a taxi).
Your servers then action the booking, and return the taxi booking confirmation response back to the the user, in real-time.
By outsourcing the language processing part of your logic, you can focus on your core business: the end-to-end user message flow, and user experience.
Natural Language Processing (N.L.P) is tricker than it seems.
No doubt you've already hit one or more of these problems parsing a 'simple' sentence:
' [I want to book a] taxi 10am tomorrow [from the] main plaza. '
' I want to bok a taxu 10am tomorrow from the main plaza. '
' Main plaza, 10am tomorrow, book taxi. '
' Please book me a fun, cute taxi driver at 10am tomorrow from the main plaza '
' Please (Order|Book|Reserve) me a taxi at midday tomorrow from the main plaza. '
' Please book me a taxi at (midday|12am|12h) from the main plaza (tomorrow|26th June|this Tuesday) '
' Do you have (any taxis|a taxi) at 10am tomorrow from the main plaza '
' Plz book taxi at 10am tmrw from the main plaza please. '
Learn a new language suited to N.L.P, then solve the challenges above.
Either learn AI/ML in-house, or recruit / acquire experts.
Immediately outsource the issue and make progress.
Do the first 2 solutions fit into your Agile process in delivering something useful and useable soon ?
Why not try us as a technology partner right now, and then start ramping-up solutions 1 or 2, so you're covered over a longer term?
You focus on the delivering the customer-facing solution, we'll work in parallel handling language processing.
We want to onboard you quickly, get the integration done, and deliver something quickly.
We assign you an already pre-configured, dedicated, secure server and give you the domain, port and request url.
On receiving requests, we return pre-canned / stub json or xml responses, for sample 'ok' and 'error' responses.
This gives you an HTTPS integration point in the cloud, so you can immediately start firing messages at us, and handling the responses.
This will be done within 48 hours.
Optionally, we can include any extra security on top of TLS - including signatures, or public-private keys.
A brief discussion about the message flow, types of questions, phrases etc. you need to parse.
We document this as set of initial 'conversation test cases' that shape the first deliverable.
Here we define, for a given set of phrases, how your lingo munger should respond in each case.
From the conversation test cases, we agree on the first case to implement eg. booking a taxi, then do it.
As soon as the lingo munger is doing 'useful work', we'll give you a new, versioned Test Server url to fire real requests at.
You may want us to go deeper within the grammar of a single use case, or go wider across different use cases - it's up to you.
We're used to dealing with Agile approaches, so we can turn on a dime to suit where you are for any given sprint.
We're part of your team, so our jointly agreed sprint goals (conversation test cases) define our deliverables to you on a given date.
Your fees are based on initial Onboarding, Consultancy hours, Message traffic and Service Provision.
You pay an up-front $450 fee per project, to get you up and going, which includes the work done up to, and including Onboarding above.
After Onboarding, the hourly rate for time spent on your project is $50 per hour.
Service Provision & Hosting is billed by the number of Test and Production servers you reserve / use.
|Type||Service||Price Per Month|
|1||Test (VPS)||$100 per month|
|2||Production (Dedicated - Bare Metal)||$500 per month|
Test Servers are limited to 5GB bandwidth per month, then a Production server will be needed.
Compared to Test Servers, Production servers have more memory, processors and processor cores.
So they are faster, and can handle more concurrency / transactions per second at higher loads.
Onboarding and integration message traffic is free of charge.
As soon as the lingo munger is doing 'useful work', and responding on your Test Server url, you will have access to a web interface showing all requests and responses, which is the basis of the Message Traffic Fees.
Message traffic is billed by: Message Qty (1 HTTPS request/response pair) x Message Type.
|-||Insecure / 80||Never available - port disabled|
|1||TLS (default)||$0.005c per request/response pair|
|2||PKI Signed||$0.007c per request/response pair|
|3||PKI Encrypted - 2048 bit||$0.010c per request/response pair|
PKI-related message bodies use more bandwidth and processing power, so have a higher message fee.
On a per-request basis, you can use a url that determines the message type, on both Test and Production servers.
Prices valid for 2018 / Q1 & Q2 2019.
We'll be in touch within 24 hours.
(+34) 625 297 181
ES: (+34) 625 297 181
US: (646) 905-0016
BuzÃ³n 253, C/Patricio Ferrandiz 40,
03700 Denia, Alicante, Spain.