COMING SOON! - Join the free early access programme

Realtime question answering

Powered by artificial intelligence

Search on steroids. Product Sage doesn't just return matching documents or snippets, it actually answers questions directly.

Benefits

Benefits of question answering

Reduced operational costs, improved conversion rates and an amazing user experience

24/7 Self service

Product Sage enables customers and other stakeholders to get instant answers to their questions without resorting to a knowledge base or contacting your team.

Inline answers

Improve your bounce rates by keeping visitors on your landing pages. Product Sage serves visitors on the landing pages themselves. No more Googling for answers!

Automatic FAQs

FAQs are a powerful usability and SEO tool, but writing product specific FAQs is time-consuming. Product Sage can automatically generate product specific, SEO compliant FAQs.

Self learning

Our machine learning allows Product Sage to continually improve with little or no oversight or development effort, freeing up your team for more important tasks.

Features

Developer features

We've taken some really advanced concepts and made them accessible to all

Question answering

Extractive and generative question answering using custom transformer based natural language models

Product Sage will automatically return generative answers in cases where it will result in a better user experience.

Alternatively you have the flexibility to explicitly request extractive answers.

Pre-trained language models

Product Sage includes pre-trained, domain specific natural language models covering different scenarios including:

  • Product & service questions
  • Shipping & returns
  • Terms and conditions

We continue to add new models based on developer feedback.

Semantic search

Many users don't have questions, but instead want to drill down on specific content e.g. shipping rates.

Product Sage supports traditional keyword search but uses its powerful natural language processing to deliver more targeted search snippets and document matches.

Automatic model selection

High accuracy is achieved by selecting different models for specific scenarios e.g. product vs shipping questions.

The best model is selected by taking into account the page itself (product, category, terms etc) along with the nature of the question.

Self learning

Our natural language models are continually tuned and refined, resulting in improved accuracy and a better user experience.

In addition to the tuning performed by our own team, Product Sage learns and improves based on user and agent feedback.

Multiple data formats

Product Sage can ingest a range of document types including:

  • HTML
  • Plain text
  • PDF
  • Microsoft Word (Docx)
How it works

Extractive vs open generative question answering

Extractive answering

Extractive answering extracts verbatim text from a document in response to a question e.g.

Question: Do you offer next day shipping?

Document source: ... We offer next day shipping ...

Answer: We offer next day shipping

Generative answering

Open generative question answering generates an answer, using a document for context e.g.

Question: Do you offer next day shipping?

Document source: ... We offer next day shipping ...

Answer: Yes

Screenshots

Dialog widget

Simple popup style dialog box. Best suited to simple flows

How it works

Data ingestion & query handling

Product Sage employs a form of artificial intelligence known as extractive and generative question answering - a two-step process.

1

Document ingestion

Documents are ingested by our web crawler or through the API and placed in a redundant, persistent queue

2

Parsing & pre-processing

The documents are split into smaller chunks, then pre-processed by our natural language processing models

3

Indexing

The chunks of text are stored in an optimised, tree like structure for fast retrieval at query time

4

Query parsing

The query is processed by a neural semantic parser (NSP) to understand it's true semantic meaning

5

Document retrieval

Candidate documents are retrieved and ranked according to their relevance to the semantic query

6

Answer retrieval

The semantic query in its logical form is applied to the candidate documents. The best answer is selected

FAQs

Frequently Asked Questions

Learn more about Product Sage and how it compares to competitor products

  • Can Product Sage be used as an employee tool?
    Yes absolutely! whilst our Javascript widgets are intended for use on a public website our API can be used to quickly build in house applications. We're also working on a new front end for internal, employee use.
  • Do you offer a free open source version?
    No. The real value of Product Sage lies not just in the software but the data, in particular the realtime feedback from our end users and agents. A cloud based solution allows us to collect data in realtime, use it to improve the models and propagate the updates to all clients.
    Additionally, running an A.I. platform in production is both difficult and expensive (hint: you need specialised GPU instances). A cloud based solution allows you to offload this to us.
    We do offer free plans, so Product Sage should be accessible to all.
  • Can I use the developer plan for production?
    Yes, you're welcome to use the free developer plan in production, subject to the quotas. Be aware that unlike the paid editions, the developer plan comes with no warranty, SLA or indemnity.
  • Is Product Sage ready for production?
    We're currently in developer preview. We expect to move to general availability during Q3 2022. Now is a great time to get involved and share your feedback and feature requests!
Pricing

A price plan for everyone

Plans for solo developers through to enterprises

Developer (EAP)

For developers working on POCs, MVPs, Solo and other small scale projects.

  • All core features

  • 5 document index

  • 100 queries per day

  • Free forever

Developer access

Free access to the core platform

Forever free

Not for profit

Free for open source projects, not for profit organisations and registered charities.

  • Full platform features

  • 10,000 document index

  • Unlimited queries *

  • Rolling 30 day contract

* Subject to our fair use policy

Not for profit

Reserve your free account now

Free Reserve

Forever free

Professional

Perfect for SMBs & startups. Our professional plan offers predictable flat rate pricing.

  • Full platform features

  • 10,000 document index

  • Unlimited queries *

  • Rolling 30 day contract

* Subject to our fair use policy

$499 pcm

Reserve your professional account now

$249 pcm Reserve

30 day commitment

Enterprise

For organisations requiring custom models, advanced security, audit logs and compliance tools.

  • Custom models

  • AWS VPC peering

  • Private data silos

  • Granular access controls

  • Audit logs

  • Support & consultancy

Contact us

Speak to a member of the team

Quote me Contact us

12 month commitment