Ubiquity Changelog

Week of Aug 17, 2026

• Continues vector-db data migrations
• Data egress integration for large social media monitoring platforms

Week of Aug 10, 2026

• Migrate vector db connections to use read/write aliases for more complex underlying data migration strategies
• Begin migration to newer vector store optimizations, with better quantization, better memory handling policies and cleaner system for handling large amounts of data ingest

Week of Aug 3, 2026

• Cleanup logic for setting chats private / not private
• Admin panel usage details for tracking system usage without the older grafana system
• Add data used for cost tracking and analysis
• Front end panel for adjust permissions on a private chat, choosing users to access the chat, reaction to certain settings that are not permitted with current logic
• Scaling changes for vector-db

Week of Jul 27, 2026

• More social media coverage

Week of Jul 20, 2026

• Develop internal proxy routing system in go. Able to redirect and retry proxy connection until a proxy accepts the attempt
• Increase coverage of alternative social media networks
• Increase memory limits for basic data crawler
• Set basic views of data series to last 30 days by default, to prevent long load times for larger datasets where older data is not of interest
• DB migrations to prepare for newer private chat settings
• initial code introduction for handling private chats

Week of Jul 13, 2026

• Bug fixes for multi-location queries
• Apply retry logic in search systems
• Update internal subsystem sdk and make changes to accommodate the update
• UX cleanup on source counts to accommodate for pagination changes
• Add authors as a filtering field for the streaming endpoint
• inherit data set visibility from pipeline permissions
• add path for reporting agent to reference data sets

Week of Jul 6, 2026

• Refactor common url transformation code into a specialized library for easier reuse across code bases
• Implementation of behavioral data subsystems for tracking major non-western statistical governmental datasets
• tweaks to agent handling of social media sources
• handle token timeouts where AWS library fails in python lib for bedrock mantle
• Add social media handlers for more networks
• Fixes for handling more non-english language search
• Work around for un-even proxy and data provider coverage
• Rework site availability / reachability testing
• update underlying infra/libraries/services
• Better retry logic

Week of Jun 29, 2026

• Update handling of 0 value refresh_interval to transform to nil value, since a refresh of 0 implies constant refreshing… and causes issues with JSON update api

Week of Jun 22, 2026

• Creation of context report builder… as subsystem to builds reports about the target topic and location to increase agentic awareness of the information environment prior to discovery processes

Week of Jun 15, 2026

• Discovery Process now has history and ability to re-open the chat interface to make refinements to pipeline creation details (topics, locations, etc)
• Pub / Sub system updates

• DB migrations for newer discovery chat sessions capability
• User Invite system for admin panel, including email
• Re brand login screen for clearer consistency

Week of Jun 8, 2026

• Administration panel for front end: ability to natively manage users, organizations, running jobs, etc
• Clean up old suggestion system and change labels to "create pipeline"
• Fixes for CSV downloader to align with new pagination system

Week of Jun 1, 2026

• Handle paths to online datasets better, not rejecting them as individual artifacts
• Add new pagination controls for number of sources
• adjust the front end app to use newer filtering system
• stylistic adjustments

Week of May 25, 2026

• Add in options for better pagination control for pipelines
• migrate api to a more functional new endpoint for viewing the entities for a feed, including better pagination controls

Week of May 18, 2026

• allow display of document scores in data browser module… showing the scores of document matches and scores of metrics that the user has requested
• vector-db index adjustments
• pull refresh data into pipeline list api
• adjust sort order of behavioral data
• adjust memory budgets

Week of May 11, 2026

• New behavioral data feature: migrations, agents updated for location and ingesting behavioral data and displaying it to the end user.
• ability to upload behavioral data directly
• Better handling of unrouted pipelines
• Searched vectors can return either full payload of just UUID
• update postgres extensions
• Clean up report of replicas, adjust pstgres params
• Improvements to data upload ui
• ability to parse xlsx files
• various ui formatting fixes for clarity and style

Week of May 4, 2026

• Improvement to pipeline router ui
• UX improvements for finding and linking pipelines in the router
• Lower cost model for high usage requirements of metrics generation
• Make metrics available to databrowser module on front end
• add separate vector-db index for uuid, so we can use those for artifact search instead of document_ids

Week of Apr 27, 2026

• Increase social media capabilities
• improvements to handling of tenancy fields in vector-db
• better handling of video transcripts
• CI/CD memory requirement adjustments
• updates to pipeline router to make it easier to use
• test enricher for data generation
• update session storage so chat can reference internal history

Week of Apr 20, 2026

• Feed API updates and vector db optimizations, adding index to commonly needed column
• adjust memory usage on primary crawler container, better handling of memory overuse conditions

Week of Apr 13, 2026

• Memory adjustment for agents
• Adjust weight of local sources in evaluation agent.
• Add presets to front end agent configuration, so users can replicate well designed settings for various purposes
• increase use of parser models for agent responses
• add front end for metrics generation/enricher service
• update LLM model version
• Postgres parameter tweaking for larger workloads
• underlying infra/service updates

Week of Apr 6, 2026

• Add pipeline to pipeline routing feature: allows users to connect one pipeline to another and re-use lists of sources from and to multiple pipelines.

Week of Mar 30, 2026

• Improvements to source uploader
• Add geo coordinate associations to all artifacts and entities
• add tools for data exploration by chat bot… ability to access metrics
• various CI/CD improvements
• Make metric data available to front end via API

Week of Mar 23, 2026

• Creation of new Enricher Service, for arbitrary prompt based data extraction
• CI/CD setup for enricher
• Add more hints for topic discovery
• CI/CD improvements for postgres image update and management
• Improve handling of image data extraction
• Add Postgres extensions for time series database
• Create db for metric storage

Week of Mar 16, 2026

• Add method to download csv of chart data
• Clean up charting features to simplify interface
• LLM cost cutting improvements
• Data extraction for more types of online content
• Add feature flagging hooks to front end app
• improvements for handling imported sources
• DB schema updates for Metric Prompts Feature

Week of Mar 9, 2026

• Improvement to social media monitoring for alternative networks
• Improvements to source validation on social media channels
• Creation of entity splitter
• Set reviewed state to nill for discovery jobs, alter semantic meaning: false = review needed, true = reviewed, nil = undecided state
• Maintenance tools for entity typing

Week of Mar 2, 2026

• Increase capabilities of source upload… ability to upload multiple urls per entity
• Adjust evaluation requirements for uploaded sources, and add more user options for it.
• Adjust data chunking strategy for job parallelization
• Detect multi-entity platform sites and single entity sites

Week of Feb 23, 2026

• Implement Swagger documentation generation
• Update api limitations
• Implement automated entity typing for determining if entities are individuals or institutions

Week of Feb 16, 2026

• Create asynchronous data delivery infrastructure for data crawler, and connect to pub/sub for handling delayed data
• More automation for internal certificate factory and certificate distribution
• Improve error handling on long running jobs
• Add vector db client access to central API layer
• Use k8s reflector to help distribute secrets

Week of Feb 9, 2026

• vector db tuning, adjust params for related vector traversal
• Create front end workflow guide system… to help users walk through the pipeline creation process
• various adjustments to managing lon running pub/sub queues
• front end code management cleanup
• various front end icons for user clarity
• adjustments to prevent browser referer metadata transmission

Week of Feb 2, 2026

• Add internal systems for basic usage tracking
• vector db tuning, allow nearly consistent data display for speed improvements on front end, to allow interruption free data ingest
• Remove block vector db call from frequently used endpoint

Week of Jan 26, 2026

• Source uploading system added
• Allow uploaded sources to skip standard validation
• Add more asyncio.timeouts to extraction content crawler
• Various Library Updates
• Add account profile management to front end
• Adding dashboard to read usage data
• Various bug fixes for organizations
• Various bug fixes for social media
• Production cluster scaling improvements

Week of Jan 19, 2026

• Adjustments to Video Blog Crawler
• Adjustments to basic crawler… using asyncio.timeout to cut off long running operations
• Standardizations of nomenclature
• Present user with more Details on related entity urls
• Update CI/CD for production deployments, standardizing secrets, adjusting memory requirements
• Various bug fixes for social media analysis
• Various front end bug fixes
• UI simplifications, removing confusing information bubbles, adding hints to agents for nomenclature,
• UX adaptations: allow chat bot to set persistent filters on data display, prevent scrolling on map causing map zoom -instead of page scroll, more places to collect user feedback
• AI Model swap for performance