Cloud-based Data Platform to lead the Data Economy
Apply the power of Artificial Intelligence to your organization's most valuable resource: your data
What We Do
Iceberg's data platform allows anyone to import, crawl and connect data sources, enrich and combine datasets with machine learning modules, rules-based automations and human workflows, to generate exponential value that transforms products and entire organizations.
All in an integrated, scalable and open cloud solution.
Easily Import your data to Iceberg's cloud or crawl, parse, extract & convert data from any public or private source. Fully-scalable.
Create apps using Iceberg's GraphQL API.
Search and Browse on the web. Export enriched data to your existing systems.
Refine your Data Model
Import your data and break content silos
Define your Graph Data Model where records are interconnected via relationships, and both may be annotated. Take advantage of a Dynamic Ontology.
Bring the power of A.I. to any Organization
Enrich & Combine your data
Enrich your data and data relationships with machine learning models, rules-based automations and human workflows. Retain full Provenance control and Edit History.
Workflows provide task-oriented interfaces that optimize human data enrichment tasks
Distribute tasks to internal or external teams, review outputs and track progress
Plug-and-play integrations with ML-as-a-service providers such as IBM Watson
Explore, search and visualize your data in real-time
Data Workspaces with branches so that your editors and data-scientists can experiment with data transformations without impacting your production users
Crawl Data from Private and Public Sources. Create crawl jobs and parsers to efficiently import data from public sources
Engagement module to automatically collect and aggregate user engagement data and use it to feed your machine learning algorithms
Iceberg Text Engineering provides built-in integration with the GATE framework to enable hand-crafted data extraction pipelines from text
Probabilistic Data model with Quality Score. Quality Scores can naturally describe the uncertainty associated to ML outputs and enable editors to override the output of automations if necessary
How it works
Optimized Databases & Indexes ready for ML algorithms
Import or Connect to any Data Source. Import structured data from XML, CSV or RDF files or connect to your Databases using built-in connectors
Entity recognition and entity reconciliation support
Built-in Classification, Search Ranking and Recommendations Algorithms ready to test and apply to a selection of documents or your entire dataset
GraphQL APIs to consume Icebergs data directly from your Apps
Rule-based automations and Mass Updates allow domain experts to turn their knowledge into rules and get instant results
Fully-scalable and ready for any number of documents
A.I. Projects go faster when you can leverage vertical datasets, informed graph data models and pre-trained ML models . We are developing a network of partners to provide support to the different Industry Verticals
- A.I. powered search: learning-to-rank and search suggestion services
- 80% reduction of time and cost in Content Harvesting projects for a Global Publisher (2,000+ sources)
- Automatic Summarization and Classification of 20MM+ documents in 4 languages by extraction of Key phrases (Semantic ML model)
- Multilingual Dynamic Taxonomy based on a Semantic ML, and human supervision to normalize and reconcile entities
- Document-to-document and user-to-document recommendations based on Semantic ML model and User Behavior
- Q&A bot
- Automated Knowledge Base Construction: building a knowledge base by aggregating the facts extracted automatically from textual information and aggregated external data sources
- Training custom classification and regression models using either Iceberg native models and API integrations (IBM, Google, Azure, AWS). Results-comparison to select the best performing model
- Legal Analytics Prediction model for Spain and Canada
- Data Exploration to gain insights via visualizations (trend analysis, semantic graph)
- Prediction model for Stock and Commodities Markets connecting external data (prices, news) with internal proprietary datasets
- Pricing algorithms for a hotel chain connecting external and internal data and prediction models
- Classifieds Website: Spam detection, incomplete and duplicate advertisements detection
Iceberg is entrepreneur and blockchain-friendly. We consider equity and/or crypto tokens in exchange for platform access and consulting services for ventures with a strong team and ambitious plans based on smart applications of Data and Artificial Intelligence.
Access by Invitation
Our team is currently working with selected enterprise clients and global partners, and evaluating new opportunities on a case-by-case basis. If you feel Iceberg could be a fit for data management and AI needs, please contact us with a summary of your requirements. Thank you.