Take your Criminal and National Security Analytics to the next level using state of the art technologies even on your single laptop GPU. Leverage Digital Twin Interrogation, Automatic Fact Extraction LLM and perform Country scale Event Space analytics. Utilize unparalelled quality algorithms and AI models with 1-Click Case Resolution to reduce your workload from months to seconds with no upfront costs. Taltosh works as a standalone set of products, but also integrates in your data pipeline.
Taltosh Technology Components
TALTOSH Interrogate
Digital Twin Interrogation
Stand-alone licensable component that you can integrate in your pipeline
Create a Digital Twin from your data—whether it’s textual payload from structured event data like IPDR, CDR, GPS, transactional logs, or unstructured text from email, transcripts, docs — and perform an interrogation on it, using our advanced LLM RAG (Retrieval-Augmented Generation) technology. It means you can create personas from your scoped data (suspects, groups, places, devices) and chat with them (ask, summarize, classify)
TALTOSH FX
Fact extraction from unstructured data
Using proprietary fine-tuned, on-premise LM models on your own data
Taltosh transforms your textual data to a set of factual events, coordinates, identifiers, relations and much more, using proprietary, triplet trained LM models. Taltosh runs through your repository of pdf, docx and already stored textual data and extracts a knowledge graph, emitting a set of: events identifiers, entities, places, relations past and even future events based on intents.
TALTOSH Resolver
Resolve your case with 1-click!
Case specific algorithms that you can run on Taltosh or embed in your own pipeline
Compress days of step-by-step analytic work into 1-click. Taltosh offers A library of active algorithms that embed into your data processing pipeline as microservices. When running (automated or triggered), a Resolver performs multiple steps of analytics and inferencing on your factual dataset and potentially resolve your investigation case by emitting a set of data indispensable for decision making
TALTOSH Filters
Complete integrated suite with data storage and GUI
Utilize Taltosh features, Data Retention and manual investigation in one place
Taltosh Filters enables you to ingest, transform and visually scope your data of interest. Use explicit filters for identifiers, entities, events. It’s your one-stop-shop for utilizing the Interrogator and case Resolver modules. Taltosh visual filters let you to search on maps, discover connections within a relational graph, and investigate time series data characteristics and it’s also a key component to analyzet the decision-making grade results from the Resolvers.
Decision Grade Answers Immediately
Taltosh doesn’t generate yet another data from your data. Taltosh focuses on goal oriented results.
Taltosh Fusion
In case you need all features in one: we got you covered with our complete suite
Central GUI
The central GUI for managing operations, including workspace, case building, data ingestion and filtering, network building, running pre-defined or case-specific queries, defining autoresolvers, and exporting results.
Workflow
Taltosh workflow is a cycle of scoping your data using filters for entities/groups or phenomenons. The collected entities then can be digital twin interrogated, analyzed or can be subject of certain case resolvers
Analytics
Digital Twin Interrogation, over 40+ Case Resolvers and a host of manual Filters are here to help finish your case within minutes instead of days of manual statistical data analysis or programming
Data Ingest
Taltosh supports CSV and API streaming for live data ingestion, offering an intuitive mapping interface to align event data with your data model effortlessly. Taltosh also supports automatic factual event extraction from unstructured .txt, .doc and .pdf files
Data Retention System
The Taltosh Data Retention System (DRS) integrates high-performance, real-time databases (columnar, document, graph and file) using a distributed query architecture, ensuring data governance and secure isolation.
Automation
Apply filters when new data is ingested, automate actions taken when a filter is hit, send notifications to certain API endpoints when a malignant activity is detected
- Run Taltosh on advanced consumer hardware to quickly visualize, search and filter your cases, chat with your data and resolve impossible scenarios
- Scan through billions of datapoints, instantly recognizing patterns and verbal references – without any assistance. Just import your CSV dataset.
- Without any IT expertise, convert your complex, intertwined and entagled set of documents, pdfs and textual files into an actionable set of entities, events and relational network
- Reach analytics supremacy by installing Taltosh on a cluster data center. Ask complex questions from the Digital Twin of your data and receive summaries alongside with the facts
Do you have an urgent case?
Even if your case involves billions of GPS, cellular or other mixed data, just call us and receive a real-life demonstration and even better a case closed!
End to End Case Resolution
Provide your complex case, and we offer our vetted experts, compliant practices, software, and dedicated effort for the most efficient, immediate resolution.
Forward Deployed Engineer
With a Taltosh license, gain access to our engineers for case work, platform enhancements, custom data acquisition tools, or ML pipeline development.
NLP Development
We create an advanced, fine tuned Digital Twin of your data to chat with, utilize LangChain functions & other automation and even fine tune models to solve certain tasks within your data.
Feature Engineering
We understand the analytics requirements that solves a case for your data and implement scalable case resolver algorithms that fit in your big data processing pipeline
Custom AI model building
For investigations needing fact extraction, automation, scoring, recognition, or language processing, we support building custom AI models from scratch, including tailored training interfaces and performance metrics.
Data Engineering
We provide assistance in building ETL pipelines, feature engineering, data storage optimization and performance improvements on your big-data sets.
- Digital Twin Data Interrogation
- Automatic factual event extraction!
- One-click case resolvers!
- No data scientist required!
- Runs on your consumer hardware!
Worfklow Explanation
01
Import data
Add your CSV files for fast importing or point to a set of unstructured pdf and txt files. Taltosh will extract and store results
02
Collect targets
Visualize your data, run manual filters on your data: define the scope of your data of interest
03
Refine your case
Chat with your data, run specific case resolvers compatible with your data for 1-click resolution
04
Finalize
Export your results, present your results and create alarms to catch patterns and future data changes based on your current workspace
Do you have an unsolvable case?
Even if your case seems utterly complex we might already have a case resolver waiting for You!