Project Description

BaaS and AI crypto trading solution provider

Novi Sad, Serbia
Established in 2015

vizlore.com

Teaser of the company

Bright Habitat is VizLore Labs Foundation spinoff, with the mission to lead regional ICT technology related innovation and seed entrepreneurship culture. Bright Habitat promotes a wider usage of cryptocurrencies with its AI based cryptocurrency trading solution and lowers the blockchain technology adoption bar by providing ChainRider Blockchain as a Service solution for simplified, user-friendly and fast blockchain development.

Bright Habitat has developed the following blockchain-based prototype in the scope of BlockStart:

ChainRider Blockchain as a Service

chainrider.io

There is a scarcity of blockchain developers, and building internal competencies is expensive and time consuming. The process of building blockchain based prototypes and MVPs usually lasts a couple of months. Existing blockchain solutions are either too generic or self-centered, incomplete, and hard to integrate with existing systems.

In order to overcome these issues, we provide ChainRider Blockchain as a service solution. ChainRider BaaS drives wider adoption of blockchain technology by offering fast prototyping (minutes instead of months) and easy to use blockchain service for fintech. Through simple web forms, the end user is able to configure and deploy a blockchain solution and with a web developer knowledge integrate it with existing business workflows and applications. ChainRider comprises three main services: 1. Blockchain network generator which makes setting up a permissioned Hyperledger Fabric blockchain easier than ever before. Users can choose to deploy their network on local infrastructure or use our click-to-deploy feature to deploy the blockchain network on Google, Amazon AWS or Microsoft Azure cloud in less than 15 minutes. 2. The Smart Contract Generator service allows for the creation of Hyperledger Fabric NodeJs smart contract to be built and deployed in less than a minute. 3. Blockchain marketplace comprising advanced smart contracts and blockchain network templates which help users add more sophisticated features to their networks and build MVPs even faster.

For more information, download:
leaflet
Pitch Presentation

Technical development during Prototype stage

Through BlockStart, ChainRider BaaS has been expanded with a service named ChainRider Marketplace. The marketplace is a space for publishing templates of more sophisticated and use-case specific Hyperledger Fabric smart contract and blockchain networks. The prototype phase enabling the marketplace has included the following development milestones:

UX & UI design of the marketplace, dashboards, user management, corporate registration and new pricing model

ChainRider Python3 and NodeJS backend applications have been expanded to support marketplace workflows, dashboard, user management and corporate onboarding. This included REST APIs for: resource usage for dashboards, payments and plan management, user management and invitations, registration of corporate entities, publishing of smart contract and blockchain networks on the marketplace, creating, generating and deployment of blockchain networks for corporate users

The aforementioned features and services were also implemented on the ChainRider frontend application in Angular 8 in line with the UX and UI wireframes

19 smart contracts have been added to ChainRider marketplace, showcasing: CRUD operations, bulk CRUD operations, communication with 3rd party APIS from smart contracts, contract-to-contract communication, private-to-public blockchain interoperability, etc.

4 network templates have been added to ChainRider marketplace, showcasing real-world use-cases: energy efficiency incentivisation (with loyalty tokens), cryptocurrency micro-payments on Hyperledger Fabric, seed-to-shelf food provenance and load carrier-shipper matching and provenance

Our 1-year product roadmap:

  • Migration to Hyperledger Fabric 2.0 (Oct 2020)
  • Support for fully managed blockchain as a service solution (Dec 2020)
  • Vault management of cryptographic material (currently part of the configuration archive) (March 2021)
  • Opening marketplace for contributions by ChainRider users (June 2021)

Business development during Prototype stage

  • Bright Habitat’s pay per use business model with individual accounts was not well received by larger companies looking to have collaborative environment for their blockchain development teams and to be able to more precisely plan the budget for their blockchain projects. In order to solve the challenge, the company upgraded the ChainRider BaaS solution to support registration of companies with multiple team members with different access rights. It has also updated the payment model for registered organizations by providing multiple tiers with monthly plans depending on the amount of resources available for blockchain network hosting, thus allowing solution adapters to properly plan the budget for their blockchain projects. An administrator of the registered organization can invite other users to join the organization on ChainRider, thus sharing the organization’s resources. There are two pricing plans, Academic and Corporate, each offering multiple tears enabling flexibility of resource usage. In general, the tiers are formed based on the amount of resources available for hosting blockchain solutions on ChainRider platform, but each tier also include access to specific set of tools and services available on the platform as well as technical support

  • Bright Habitat has compiled a business brochure for ChainRider Baas and collected information about 90 companies in Hyperledger Fabric business domain. 66 invitations have been sent on LinkedIn. 35 invitations were accepted, and multiple follow-ups have been performed with them. With 4 of the invitees the company has organized a teleconference call and presented the solution. In addition to LinkedIn campaign, a cold email campaign has been run using the aforementioned business brochure for the initial contact.

  • Bright Habitat has acquired 2 paying customers during the project and 3 declarations of commitment from SMEs interested in participating in the pilot stage.  It had 9 commercial presentations of ChainRider BaaS service. At the end of the Prototype phase, they plan to hire a sales and marketing associate.

  • Multiple blockchain networks (10+), mapped to specific use cases were deployed as part of the live demo performed for potential adopters of the ChainRider BaaS solution. Bright Habitat has also published video learning materials accompanying implemented blockchain marketplace templates and included them within the ChainRider knowledge base.

Bright Habitat has developed the following blockchain-based prototype in the scope of BlockStart:

ChainRider AI-based cryptocurrency trading

Reports indicate that more than 95% of traders are losing money. The average individual investor underperforms a market index by 1.5% per year; and active traders underperforms by 6.5% annually. Among all day traders, nearly 40% trade for only one month. 80% of day traders quit within the first two years.

Relaying on individual financial indicators alone is not enough for successful trading. Manual order execution is unfavourable as it requires strict discipline, emotional distancing and often leads to errors.

In order to solve the cryptocurrency trading challenges experienced by 95% of traders, Bright Habitat has applied concepts such as artificial intelligence (AI), data analytics and automation. Its solution is based on its real-time growing data set, which currently counts close to 1 billion data points of executed trades. In order to capitalize on this data, the company employs AI which uses deep reinforcement learning to generate trading signals for 5 cryptocurrencies: Bitcoin, Ethereum, Bitcoin Cash, Litecoin and Dash. Deep reinforcement learning field of research has been able to solve a wide range of complex decision-making tasks that were previously out of reach for a machine. It is a category of machine learning where intelligent machines/agents can learn from their actions similar to the way humans learn from experience. This concept is at the core of Bright Habitat’s trading solution and it yields significant return of investment, by detecting correlation between more than 30 financial features simultaneously. The company’s solution enables automated execution of trades in customers’ accounts with its order execution engine, driven by AI-generated trading signals. Its current machine learning models provide return of investment which ranges from 0.82% to 1.95% per closed trade, depending on the portfolio spread (selected cryptocurrencies). On average, the return of investment for the first 5 months of 2020 was 101% (with included exchange trading fees of 0.15%).

For more information, download:
Leaflet
Pitch Presentation

Technical development during Prototype stage

With BlockStart, Bright Habitat has been able to optimize its deep reinforcement learning models used for generating trading signals, by adding four new financial features:

  1. Stochastic oscillator
  2. Moving Average Convergence Divergence (MACD)
  3. Hull Moving Average (HMA)
  4. Ichimoku Cloud.

The architectures of the neural network models have been changed accordingly: updated number of layers, neurons and dropout. In conclusion, the optimized models’ return of investment is on average 30-40% better than the previous models.

An automated trading service (bot) has been implemented for different exchanges and trading strategies:

  • An automated trading service (bot) for CoinBase Pro exchange has been implemented to trade with BTC, LTC, BCH and ETH. CoinBase Pro offers industry-grade, safe and secure APIs for order execution and dynamic fee structure

  • A combined trading bot capable of trading multiple cryptocurrencies from one account has been implemented for CoinBase Pro exchange. This enables trading with multiple cryptocurrencies based on one trading signal, for example: the bot tracks the trading signals generated for ETH, but reacts by trading ETH, BCH and LTC

  • An automated margin trading service (bot) for Bitmex exchange has been implemented to trade with ETH and XBT. The bots working on the Bitmex exchange have been optimized for short positions

The following product roadmap is envisioned for ChainRider AI based trading solution:

  • Implementation of trading bots for new exchanges (Binance, BitFinex, Houbi) (Sept 2020)
  • High responsive trading signal generation (from 1h to 5 minute) (Nov 2020)
  • Optimized models with new financial features combining LSTM and DRL (Dec 2020)
  • Building a web/cloud platform and providing AI trading solution to end users – B2C business model. (Feb 2021)
  • Order routing support (March 2021)
  • Automated service for strategic spreading of portfolio for high volume transactions (May 2021)

Business development during Prototype stage

  • As part of AI-based cryptocurrency trading service, Bright Habitat differentiates two business models (depending on who handles the order execution):

    • It provides trading signals and customers handle order execution: This includes signing a Revenue Share Agreement with a 20% performance fee (20% of the generated profit) charged on a quarterly basis.
    • Customers set up an account on the desired exchange (currently CoinBase and Bitmex exchanges are supported) and the company execute orders based on its trading signals. Customers decide on the portfolio distribution and share API keys that enable trading with Bright Habitat. The company sets up ML models, bots and start order execution in less than 24 hours. This includes a 2% AUM fee (annualized) and a 20% performance fee is charged on a quarterly basis.
  • Its trading service requires minimum of 200k USD initial portfolio for period of 6 months. The minimal portfolio is derived from the expected profit margin, the effort and cloud resources required to set up the AI based trading solution for a particular client. The initial portfolio from Bright Habitat’s experience is usually 2 to 5 times larger than the minimum requirement. The performance fee provided above is negotiable for portfolios larger than 1 million USD

  • As part of Brith Habitat’s business development efforts, the company has compiled a promotional video which has been made publicly available. Next, it has compiled a brochure which has been used for contacting SME adopters and potential clients. The company has collected information about 137 funds and OTC trading desks .99 invitations and 5 InMail messages have been sent on LinkedIn. 30 invitations were accepted, and of those 28 were followed up with a message or email. 26 invitees received the promotional video as part of the follow-up communication. With 6 of the invitees, the company has organized a teleconference call.  In addition to LinkedIn campaign, a cold email campaign has been run. For this campaign, the aforementioned brochure has been used

  • Bright Habitat has acquired one USA-based customer (cryptocurrency fund) which was not able to sign BlockStart’s declaration of commitment as it is a non-EU based company. It is also in the negotiation phase with a premium cryptocurrency investment fund also from the USA. The company has had 6 teleconference calls during which it presented the solution to its potential clients. The company has attended the online ANNON Summit 2020, the largest online blockchain conference in the CET time zone

  • As a result of the project, and in line with Bright Habitat’s exploitation strategy, the company has founded a new company VizLore Digital Asset Management DOO, which will primarily be focused on providing B2C solution for AI-based cryptocurrency trading

Meet the Team

Ognjen Ikovic
Ognjen IkovicCEO
Is CEO at Bright Habitat (BH) with 10+ years of experience in building innovative IT services. Ognjen is leading BH projects aimed at building AI and machine learning solutions, cloud platforms and blockchain based solutions
Milenko Tosic
Milenko Tosic CTO
Is CTO in Bright Habitat. He is a seasoned internet industry executive with 10 years of experience focused on technology innovation and research strategy
Sasa Pesic
Sasa PesicR&D Blockchain Engineer
Is an R&D Blockchain Engineer in Bright Habitat and a guest researcher/consultant at Arizona State University Blockchain Research Lab. His research interests include blockchain technology and context-aware edge computing systems
Stanimir Ilic
Stanimir Ilic Operation Manager
Is Operation Manager in Bright Habitat. He specializes in cryptocurrency trading. Stanimir is dedicated to exploring innovative business models powered by blockchain technology
Mirjana Samardzic
Mirjana Samardzic R&D data analytics engineer
Is an R&D data analytics engineer specialized in Deep reinforcement learning and LSTM
Marko Razumenic
Marko RazumenicSoftware Developer
Is a skilled software developer mainly focused on Angular frontend development

Bright Habitat’s achievements

Media features

For more information, download the company’s portfolio:

We had a great time working with the BlockStart team, and we owe great thanks to our mentor Mr. Ugnius Ramanauskas. Feedback received from mentoring sessions helped us to better position our service with respect to the market and improve market visibility. The Prototype phase was quite challenging as we had to simultaneously develop two MVPs and bring them to the level of Minimum Saleable Products though active lead generation, business development and customer acquisition

Ognjen Ikovic, CEO, Bright Habitat

Total funding received under BlockStart: €16,000 (€1,000 for Ideation Kick-off and €15,000 for Prototype stage)