We believe in robots
Our clients and projects
Our client
Fokus Portföy Yönetimi A.Åž. / Ä°stanbul
What we delivered
"Autonomous fund manager for one hedge fund"​
​
Date / Duration
March 2018 / 3 months
​​
VeriZeka’nın Finansal piyasalar ve özelde fon yönetimi için tasarladığı ‘4estRobot ©’, ilk müÅŸterisiyle buluÅŸtu.
​
Fokus Portföy Yönetimi A.Åž. Serbest Åžemsiye Fonu’na baÄŸlı olarak kurulan ‘Forest Serbest Fon’, FPR kodu ve TRYFKSP00021 ISIN numarasıyla 15 Mart’tan itibaren iÅŸlem görmeye baÅŸladı.
​
SPK tarafından lisanslanan ve denetlenen FPR serbest fonun hisse senedi portföyünü, Verizekalı'nın tasarladığı Makine ÖÄŸrenmesi temelli veri robotu yönetecek.
​
Bu serbest fon, Fokus PYÅž tarafından nitelikli olarak sınıflandırılan tüm yatırımcılara sunulabilecek.
​
FPR serbest fon ile birlikte, Verizekalı'nın teknolojisi, otonom fon yöneticisi olarak ilk kullanım alanına kavuÅŸmuÅŸ oldu.
​
Fondan pay sahibi olmak ve daha fazla bilgi için:
www.fokusportföy.com.tr
​
Fonun performansını takip etmek için bkz:
​
Our client
Deutsche Bank Securities / Romania
What we delivered
"Personal assistant for Trading Desk"
Date/Duration
June 2019 / 3 months
​​​
​​​
This project, aka ‘Eagle Eye’, signed by the client in May 2019, aimed to develop a contemporary analytical tool to support the trading decisions made by the trading desk of Deutsche Bank.
This project covered the scope of a "prototype" of the ultimate version, and, it was assumed to be the opening project of a series of projects Verizekalı and the client agreed upon.
​
The core function of the prototype was to generate trading ideas, signals or recommendations for a certain market data series (aka ticker) to the human user (i.e. the team member of the trader desk).
The prototype was designed in two layers: the back-end, and the front-end.
​
The back-end was basically designed to cater two front-end channels. The first channel was decided to be the web-browser: the user was given the ability to use the back-end from any web browser - after the authentication process.
​
The second channel has been considered as the chat-bot. This chat-bot had been designed with NLP capabilities, and it was able to communicate in an interactive manner with the user. This chat-bot was designed to provide answers to what is asked (Reactive capacity), and to alert key market turning points to the user even if s/he might not have asked any question. (Proactive capacity)
​
At the back-end, Verizekalı utilized its proprietary 4SIGN analytical engine (called Eagle Eye by the client). The back-end included a wide range of machine learning methods (mostly calibrated to time-series analysis).
​
This project helped our client:
- Enabled the trading desk to be able to follow and respond to the market developments 7/24 with the help of Eagle Eye the analytical and interactive data robot.
- Enabled the trader desk to miss no single conviction ideas and key turning in the market with the help of the proactive alerts and the messages of the chat-bot.
- Enhanced the efficiency of the human teams with the help of such machine assistants,
- Simplified the way the traders interact with the bombardment of data flowing from multiple channels.
- Shortened the trading decision cycle through faster research process and more accurate pre-trade analysis of the markets,
​
Our client
Family Office / Ankara
​​
What we delivered
"Automation of the order flow to BIST via Integration to Bloomberg's EMSX Terminal"
​
Date / Duration
September 2023 / 2 months​​
​​
Our Client
Prop-desk domiciled in Istanbul/Turkey
​
What we delivered
"Python/Kafka based EMS that manages the day orders"
​​
Date
November 2023 / 3 months
​​
Deliverables
We delivered to our client one light-OMS that:
​​
-
- Orchestrates the order flow of the desk,
-
- Works autonomously TWAP and VWAP orders over the day,
-
- Automates the management of the day orders,
-
- Provides a dashboard to monitor the executions of child and parent orders.
​
Our client: AGRI3
AGRI3 Fund aims to mobilise $1 billion of financing , by providing credit enhancement and technical assistance , to investment projects and businesses which protect and restore forests, promote sustainable agriculture and support rural livelihoods. The Fund provides guarantees to financial institutions (FIs) which provide credit to support eligible projects and businesses.
What we delivered
"Python/Excel dashboard for the fund to monitor and project financials, portfolio metrics and sensitivities"
​
Date
December 2024 / 3 months
​​​​
Deliverables
We designed and delivered an Excel-fronted, Python-powered financial modeling platform for AGRI3, enabling quarterly projections across cash flow, balance sheet, tranches, and key performance metrics.
The solution includes a fully automated waterfall engine, integrated reporting suite, and advanced analytical capabilities such as scenario, sensitivity, and stress testing.
It allows seamless alignment with investor, regulatory, and rating reports while giving management a flexible, high-level input framework for strategic planning and decision-making.
​​
Our client
Strateji Portföy Yönetimi A.Åž.
​
What we delivered
"Autonomous fund manager for one hedge fund"
​​
Date / Duration
November 2018 / 6 months​
​​​
VeriZeka’nın Finansal piyasalar ve özelde fon yönetimi için tasarladığı ‘4estRobot ©’, yeni müÅŸterisiyle buluÅŸtu.
​
Strateji Menkul DeÄŸerler A.Åž. Serbest Åžemsiye Fonu’na devredilen ‘Forest Serbest Fon’, FPR kodu ve TRYFKSP00021 ISIN numarasıyla, iÅŸlem görmeye devam ederken, fonun VIOP ve diÄŸer hisse senedi kontratlarını Verizekalı'ın saÄŸladığı yapay zeka uygulaması yönetecek.
Bu serbest fon, Strateji PYÅž tarafından nitelikli olarak sınıflandırılan tüm yatırımcılara sunulabilecek.
​
Fonun performansını takip etmek için bkz:
​
Our client
4Play Dijital Ä°ÅŸler A.Åž.
​
What we delivered
"Pricing and marketing engine"​
Date / Duration
February 2020 / 2 months
​​
​​​
The Problem
The revenues of the mobile operators including Turkcell are at their mature stage where by the only path to generate new growth passes from data-related services. Unsurprisingly, Turkcell prioritizes the data revenues vs. the voice revenues into 2019.
Turkcell initiated BIP platform for this purpose. BIP platform, that provide chat, contents and value added services to the subscribers, acts as one big data-revenue source for the operator.
4P Dijital Ä°ÅŸler, a partner and a member of the Turkcell's business ecosystem, assumed the responsibility to develop and manage some services provided through within BIP platform.
The Solution
(Key theme: Machine learning)
Verizekalı was appointed for the task of increasing the revenue potential of one key service of this platform, "Kim 1 GB ister, which was managed by 4P on behalf of Turkcell.
Verizekalı designed, coded, and delivered one ML applicaiton for 4Play that aimed to:
-
Collect data
-
Normalize data
-
Impute the data
-
Feature selection
-
And recommend actionable marketing and pricing strategies for 4P for this service.
​
The features
With the help of this ML application, 4P was able to:
-
Create sub segments of the user base
-
Segment the users based on the usage statistics, Personalize price and campaign offerings
-
Optimize the pricing and the timing of promotion notificiations sent to circa 6 millions users.
KPIs
The revenue of the K1GBİ service simply doubled in three months - thanks to Verizekalı's ML application that segmented the clients, customized the marketing mix to the cilent, and, optimized the promotion frequency and the timing per client.
Our client: Reitera
Family Office with AUM around EUR 30mn
​
What we delivered
"End-to-end Machine Learning Pipeline to automate the idea generation process"
​​
Date
December 2023 / 12 months
​​​​​
Deliverables
We designed and implemented an end-to-end machine learning pipeline for Reitera that seamlessly orchestrates data ingestion, cleaning, feature engineering, and model execution.
The system integrates portfolio optimization algorithms with dynamic market inputs to generate robust, data-driven, alpha-seeking active allocation decisions.
It culminates in automated order generation, and eventually enabling a fully systematic and scalable investment workflow from raw data to trade execution.
​
The project included Rust backend - coupled with Python.