Current & Recent Projects
- Virtualization – step 1: discovery of original content and creating a database of indices for all frames of interest in a video, that serves as a basis for later (re)construction of a video
- Virtualization – step 2: creation of virtual videos, which are size-wise a fraction of a traditional video file ( pointing back to data in the original file “by reference” rather than transcoding it thus creating a copy “by value”)
THE STARTING POINT
An innovative market intelligence company providing quantitative insights into the elements of commodities trading supply chain.
EMF measurements in the vicinity of pipelines’ pumping stations and infra-red imagery of oil storage facilities and refineries provides invaluable insights for commodities traders and markets forecasters.
WHAT WAS THE PROBLEM?
Unique highly valuable to the markets data sets are very capital intensive to produce and extend thus making it very expensive to grow the business.
WHAT WAS THE SOLUTION?
DistCoTech suggested an innovative way to leverage existing data sets through amalgamating them into an integrated supply chain model that can be used for macroeconomic forecasting.
We developed an extensible supply chain model as a platform to integrate time series data sets, updated in near real time, quantifying state of various elements of the North American supply chain for crude oil and refined products (pipelines, storage, refineries, tankers).
We further extended the platform to integrate publicly available macroeconomic and pricing data sets relevant to the target market.
We productized the solution by designing UX/UI and integrating it into the core user portal.
TECHNOLOGIES, OPERATIONS AND THE OUTCOME
Cloud-based solution based on a graph database modeling North American commodities supply chain, including ports and support for vessel tracking data.
Public API to support integration of the relevant 3rd-party data sets.
GEEK SPEAK BUZZ WORDS
.Net. Angular, Neo4J, R, Python, Azure ML
THE STARTING POINT
P2P marketplace with support for multi-media interactions between the counterparties.
WHAT WAS THE PROBLEM?
Inefficient platform implementation based on outdated technologies and informal adviced engeneering practices and processes made it nearly impossible to implement new features with a reasonable amount of effort. Associated cost of ownership rose exponentially and quality of development plummeted due to unexpected side effects induced by even small changes to the point of the project having been paralyzed.
WHAT WAS THE SOLUTION?
Application has been completely redesigned and fundamentally reimplemented based on Spring framework, subsequently migrated to AWS with CI/CD being implemented at the same time. Added Adwords/Facebook integration for conversion tracking.
TECHNOLOGIES, OPERATIONS AND THE OUTCOME
Integrated billing and payment solutions (Stripe, Payment Express PXPay, Xero), and VoIP call flows (Asterisk, Twilio). Implemented Gitflow workflow process with CI/CD and zero-downtime deployments to AWS.
GEEK SPEAK BUZZ WORDS
Java, NodeJS, AngularJS, Spring, Nginx, AWS
THE STARTING POINT
Linius Technologies (ASX:LNU) held a patent for digital video virtualization but lacked expertise to execute on its marketing strategy.
Substantial amount of initial funding went into failed development effort.
WHAT WAS THE PROBLEM?
non-workable non-scalable source code, failure to capture commercial potential of IP stemming from the patent.
WHAT WAS THE SOLUTION?
DistCoTech developed from scratch Linius Video Services SaaS capable of Virtualizing a video file.. Virtualization process is codec agnostic but container specific. Video file with any codec with either MP4 or HLS container can be virtualized
Virtual video can be delivered as an HLS stream with ABR or as a regular MP4 file
TECHNOLOGIES, OPERATIONS AND THE OUTCOME
Charging and Billing solution based on Chargify and Braintree Virtualized video can be further enhanced with AI data using Microsoft Video Indexer.
Metadata gathered from MSVI or any other Metadata Enrichment provider are stored for late Search in Elastic Search.
Virtual videos are a fraction of the size of a traditional video file.
Network-aware video delivery via HLS streamingScalable platform architecture capable of adopting to any network traffic requirements Commercially successful platform, currently providing Enrich and Search capability and customized videos to customers in news (e.g. Newstag), sports (e.g. Oklahoma State University) and entertainment (e.g. racing.com).
GEEK SPEAK BUZZ WORDS
Java 8, Jenkins 2, Ansible, Terraform, AWS, Azure, CI and zero downtime deployments. Distributed MongoDB replica clusters. ELK cluster, Grafana.
THE STARTING POINT 
Software platform aimed at helping customers optimize their marketing spend by tracking what is trending on the Internet in real time.
WHAT WAS THE PROBLEM?
Boost marketing mobility campaign in an automated way through different marketing channels.
WHAT WAS THE SOLUTION?
Processing of large volumes of data in near real time. High throughput rules engine: search for predefined conditions (keywords) and execute action(s) when a condition set is met.
Data for analysis is coming from various sources (e.g. news feeds' subtitles).
TECHNOLOGIES, OPERATIONS AND THE OUTCOME
Google AdWords trading platform.
GEEK SPEAK BUZZ WORDS
Java, Kafka, Couchbase, Docker, AWS ECS, Cloudformation, Troposphere.
THE STARTNG POINT
OTRO is a global digital fan club created by the world’s top footballers for their fans. This is where our founding players, including the likes of Leo Messi, Neymar Jr. and David Beckham, invite their fans to see the other side of their lives through exclusive multimedia content. Otro is a club where players and fans interact through comments, quizzes, competitions and live Q&As.
WHAT WAS THE PROBLEM?
Solution needs to be robust and scalable, particularly with respect to concurrency, to account for the sheer size of its target demographics: the total number of fans (followers on social networks) for the world's top footballers exceeds 1 billion people. Example of capacity and throughput requirements: support new user registrations on the order of thousands/min).
WHAT WAS THE SOLUTION?
Cloud-based distributed asynchronous system optimized for high end-to-end throughput.
TECHNOLOGIES, OPERATIONS AND THE OUTCOME
As soon as David Beckham announced launch of Otro on Twitter - approximately 2 hours ahead of the scheduled go-live - more than 225,000 people successfully registred in less than 20 minutes.
GEEK SPEAK BUZZ WORDS
Terraform, Kubernetes, Helm, Java, Spring Boot, Apache Kafka, Vault, CircleCI, GCP.