We are actively working on the following:
Improving dashboards / report builder, among other we are adding individual text variables to reports and reporting
Next iteration on the AI Model for KPI extraction from text, focus on named entity linking
Improving workflows related to newsletters forwarded to Midesk e-mail address
Midesk will fetch a defined API on a periodic basis, structure its content it and save it into its data repositories.
This is very helpful for passively generating market figures from SimilarWeb, Moody's Analytics, OpenCorporates, OECD, etc...
The implemented API not only extract given KPIs but also do "data acrobatics", including functions like counting (sum, min, max, mode, average) and filtering (firstWhere, unique, last, first). More function will be implemented based on our clients' user cases.
Midesk Report Builder is a drag & drop editor that connects all parts of Midesk into management-ready dashboards that are automatically updated, and exportable in the PowerPoint format.
The dashboards are based on the latest state of the "data repositories" which consolidate and manage data from various places, including via a Midesk Crawler, INBOX email forwarding, or manual entry.
The report builder is powered by a powerful tree-calculation method that can consolidate daily, weekly, quarterly figures into the right results for the client. It can also calculate and provide market sizes from unstructured figures (e.g. the total addressable Job-board market in Germany).
Midesk report builder features a multi-column resizable layout and movable blocks.
Midesk report builder offers a very flexible graphing library that aligns the web editor to its counterpart graphs in PowerPoint (real native PowerPoint graphs representation)
Midesk report builder offers "rolling periods" (e.g. last 2 weeks) so the client doesn't need to touch their graphs.
Removed Midesk website from the Product Backend and Transferred It to Its Own Container
Added screenshots to the Midesk Crawler
Added crawling email notifications: aka get informed if your competitors do something
Added crawling counted fetching of KPIs: aka get to KPIs when they are not available
Added logos to data models
Optimized data calculations for missing values
Fixed a bug that prevented registrations from going forward
Adds click-to-edit capabilities to the report builder
Adds rolling periods to calculations (aka give me the last 10 days), extending the fixed method
Adds a relative graph
Improves data cell loading
Better handling of preview screenshots generated during the crawling
Improves the News KPI annotation interfaces
Better fetching of articles from Feedly newsletters
Adds automatic KPI Extraction from websites
Fixes a bug when changing data types
Automatically fetched the information about a source when a data point source is a url or existing article in the system
Big Report Builder Overhaul
Standard reports based on a modular drag&drop online builder with blocks like text, images, columns or graphs
Automatically generated - can be based on data from other parts of Midesk, like the automatic website scraper
Has a capability to select the most interesting data to showcase
Added a nice front slide
Customisable corporate identity (colors, logos)
Embedded graphs as Excels with underlying data, not a photo
Backup slides with all data sources and links to the users' Midesk workspace
Adds JSON crawler
Automatically extract links from comments
Tracks previous and next cell for better data management
Adds a possibility to have multiple version of a data point in time and across reliability (like from news, annual reports, etc.)
Updates to PHP 7.4
As our new website has just been launched and we have just started publishing our changelog, it's a bit of an overhead to dig deep the details from the past. We will add the missing months at a later point in time.