FoundationX
Start building your path with AI
What is FoundationX?
A unified platform that helps you unleash the power of your documents in a way unseen untill AI

Advanced User System

That lets you create any complex organisational structure that your company may have

Multiple data input sources

Easily bring your data into FoundationX and keep the sources up-to-date with unique recurrence system

Flexible talk-to-documents chat

Use your documents for any task imaginable such as like chatbots, onboarding new staff or assisting sales agents

Extract structured data from docs

It’s now easier than ever to extract data that you need in a structured way using our default tags or even creating new ones

And even more in the future like:

On-premise infrastructure

FoundationX is designed with on-premise functionality in mind. It can be utilised either as a cloud SaaS platform or installed locally within your own DataCenter, ensuring complete privacy. Additionally, for every tool we develop, we integrate a local open-source AI option.

Auditing your company data & processes

Collecting feedback from your employees and structuring their responses is now faster than ever. This allows you to analyse the bottlenecks they encounter or identify areas most suitable for automation using AI solutions.

How it’s FoundationX organised?
Our architecture is planned to be secure, flexible, very fast, deployable and expandable

Advanced User System

Built on top of Organisations, Teams and Users

Multiple data input sources

Like documents, cloud platforms, scrapers and even multimedia platforms

Services

Our developments areas: Chatbots, NER, Auditing

Projects

Integrating users, data, and services into a unified unit

And even more like:

Cloud / Local AI Models

API Keys Owned by FoundationX / Bring your own keys

Various data outputs

Advanced User System

Organisations:

  • An organisation has at least one team
  • An organisation can have many teams

Teams:

  • A team has at most one organisation
  • A team has at least one user
  • A team can have many users

Users:

  • A user has at least one organisation
  • A user has at least one team
  • A user can have many organisations
  • A user can have many teams
Multiple data input sources
  • They belong to the organisation
  • All of the input sources transforms the input to text
  • Multiple input sources like: audio, video, documents, even PDFs
  • Can be reusable in any project
  • Can be kept up-to-date with one checkbox
We are also planning to support images in the near future
Services
  • Chatbots enable communication with uploaded documents through diverse inputs such as webpages, messaging platforms, or APIs.
  • Name Entity Recognition involves identifying and extracting multiple entities, including persons, organisations, and custom entities within organisational data.
  • Auditing entails conducting feedback forms, which are then processed with AI to identify various issues.
Additionally, we're investigating the addition of a service to correlate different uploaded documents with various parameters. For instance, determining if a scanned document contains an image of a car or a cat.
Projects
  • The base unit of FoundationX
  • Can use only one service
  • Can have access only to the shared documents and data provided by the creator
  • The users that can access / change configuration must have permissions
  • Can use only one AI model at a time
  • Can be cloned
  • Has at least one output
  • Has own settings, different from the global ones
FoundationX
Input Sources
FoundationX Input Sources
Our primary objective is to enable effortless importing and keeping data sources up-to-date with just one click
In progress

Documents

All sort of documents starting from Word, Texts, PDFs (including OCR), PowerPoint and any files containing texts

In progress

Websites

Importing texts and documents from any website that you can think of, well, almost lets not scrape Amazon

In progress

Cloud media platforms

Youtube / Vimeo / TikTok. Create a chat that can answer to your customers using responses from your videos

In progress

Cloud storage platforms

Dropbox / OneDrive / Google Drive / Sharepoint. Easily import data from your company cloud storage and start chatting

Planned
In the future
In the future
In the future
Planned
Planned
Cloud and Web sources will have a sync option included
Importing data from Documents
Main source for data in FoundationX, will consists of text documents
We currently support or plan to support the followings

Any type of document that contains text directly: .txt, .md, .htmlDone

Office documents from Microsoft or other providers like Word, PowerPoint or NotesDone

PDF documents even those that require OCR processingDone

CSV or XLS files that contains data in a structured wayIn progress

Importing data from Cloud media platforms
Media content platforms serve as a vital source of data for analysis and bot functionality imports
We currently support
Youtube
Done
Vimeo
Done
TikTok
Done
Features that are currently available or in the pipeline

Handling channels and extracting either entire videos or specific segmentsIn progress

Flexible limits on the number of videos that can be processedPlanned

Capability to delete videos from the extracted collectionIn progress

Capability to ensure that the sources in FoundationX remain up-to-date with newly posted videos on the channelsPlanned

Capability to extract text from either audio or video sourcesDone

Importing data from Websites
Importing data from websites was never easier. Import texts, code or files
We currently support or plan to support the followings

Scraper that extracts texts, code or files from webpagesIn progress

Flexible limits of pages and documents that can be importedPlanned

Capability to delete any page from the extracted dataIn progress

Capability to ensure that the sources in FoundationX remain up-to-date with newly added pages or infosPlanned

Scanning based on the sitemap or notIn progress

Personalized or custom scraper built with AI based on your needsIn the future

FoundationX
AI Chatbots
What are AI chatbots?
A modern approach enabling interaction with documents and unstructured data through conversations

Multiple data sources

You have the ability to interact and engage in conversation with all forms of your data, including documents, audio, and video

Multiple outputs / integrations

Integrate your chatbot seamlessly with your requirements, be it your website, messaging platform, CRM system, or automation tool

Privacy & security

Your data is hosted on our servers, which are GDPR compliant and undergo regular security testing

Cloud / Local AI Models

You have the option to select from a diverse range of models available in the cloud, such as GPT-3.5 or GPT-4, as well as local ones like OpenChat or Llama

Personalisation

You have the capability to customise the appearance and user interface of the chatbot to align with the style of your website or platform

Up-to-date

By simply checking a single checkbox, your chatbot can remain synchronized with your data sources, allowing for automatic or manual retraining as needed

Cloud and Web sources will have a sync option included
How can you use such a wonder?
This technology offers a multitude of use cases that are limited only by your imagination and the specific needs of your industry

Virtual Assistants

Experience swift query resolutions, available 24/7, with personalised assistance tailored to your customers' needs.

Onboarding Flows

Are you seeking assistance in facilitating the learning of internal procedures for a new employee?

Data Analysis

Extract relevant insights from datasets and unstructured documents, audio and video.Easily make data-driven decisions.

What kind of integrations are there available?
Our goal is to ensure our product's availability across all platforms prioritising ease of access.
FoundationX
NER - Extract structured data from docs
What is NER?
Name Entity Recognition - identify and categorise key pieces of information in unstructured text
How NER Works? At the heart of any NER model is a two step process:
1.
Detect a named entity
2.
Categorize the entity
Examples of common entities categories:

Person

E.g., Elvis Presley, Audrey Hepburn, David Beckham

Organisation

E.g., Google, Mastercard, University of Oxford

Location

E.g., Trafalgar Square, MoMA, Machu Picchu

Work of art

E.g., Hamlet, Snow White, Foundation.ai

What is our NER objective?
Our primary objective is to empower you to effortlessly define your entities and relationships while automating your pipelines with ease
What tools will FoundationX provide to enable you to achieve this? This is an area of great pride for us, and we dedicate considerable effort to ensuring that these tools are of the highest quality and produce the best results.
Public Interface
  • Only file by file upload
  • A few default selected entities
  • Our fastest model
ner.foundationx.ai
Private Interface
Train Interface