Bulent Egrilmez

A telecommunications operator needed an AI-based optimization model for the location of its retail stores. I owned the team’s deployment and created the model using Keras and TensorFlow for regression.

I managed the entire program for GDPR compliance analysis and implementation and a related data governance effort.

The client demanded a solution for the real-time tracking of quality management metrics in an immutable environment. I deployed Hyperledger Sawtooth that was also in line with their front-end IoT sensors.

The client demanded an end-to-end service design for a digitized services kiosk terminal experience. I delivered the solution using service design methods.

The client requested a solution for a layout-based classification on unstructured data of documents. I managed the team developing an OpenCV solution for that.

A major bank in Turkey requested that I revamped its loyalty program. I proposed a new concept and subsequently led a team of developers and designers on its implementation—successfully delivering the application on time and within budget.

The client requested a mobile app for its own new video-on-demand service. I managed the team and conducted service design workshops for the experience with users as well as the management of the design team.

The work I did for this project included data classification, anonymization, and the management of legal analysis for article 6.

I led the team of designers and developers to deliver a full-fledged security and venue management solution for a national football association.

The platform is based on robots that are running on MetaTrader and further versions will include a field-programmable gate array (FPGA) on demand for faster processing. This marketplace is an internal marketplace for fintech clients.

The client asked for the delivery of solutions for post-perimeter IoT and BYOD for global accounts. I managed the product development team and managed the project in the Scrum methodology. The application is based on endpoint security, mobile device management, antiphishing, and integrations to Microsoft Intune, XenMobile, AirWatch, and similar systems.

The client required a landscape transformation of their legacy systems using SAP R/3 models. I led the PMO team in an international capacity over two major time zones.

I took product ownership of the loyalty and referral platform using Scrum methodologies that harnessed several businesses of the client in real estate, hospitality, and entertainment industry in the Middle East.
The project entailed a CRM 360 system and Salesforce integration using ERC-20 cryptocurrency as a payment medium. I managed a team of 55 members in over ten countries.

I managed the product for DeFi and the underlying DAO-based network for a Blockchain company. Definition of PRDs and smart contract-driven DeFi apps. Managed the backlog and definition of the product specs and did the management of the teams.

I managed the project and defined the product for a DeFi dapp for a CEX that is focused on farming and lending of cryptos as well as further DeFi products such as NFTs and Synthetix. Was responsible for whitepaper management, roadmap management, and initiative definition.

I managed the project for an unreal-based world machine-generated, world-based play-to-earn, NFT-based game that has an IRL impact on endangered species. The application has NFT-based avatars that will be materialized in an unreal arena for PVP challenges and other quests.

I managed the landscape transformation for a major food company for their SAP transformation from R/3 to S4 Hana. I was in charge of the PMO and CR management and also helped gather the functional requirements.

I managed the transformation project for a web store configuration for 120 markets for a major FMCG company. It included Scrum-based team management, backlog management, and refining the requirements from market business units.

I designed the features and managed the technical team in the development of a real estate sales rep tracking mobile app and a dashboard for Salesforce analytics. I also used Scrum methods in technical team management.

I managed the ERP requirements—collected, refined, and prioritized them according to the SAP Business One system and manufacturing add-on specifications. I also developed the dashboard for Monday.com so the entire team could receive activity updates.

I worked on the YEK-G System, from sectors and organizations with high electricity consumption to individual users. It is designed to follow all processes of the generated electricity from the generator to the consumer in order to expand the use of renewable energy resources. It also protects the environment and makes renewable energy accessible to everyone by EXIST’s means by using blockchain technology.

This project aimed to develop a scalable, intelligent question-answering system built upon a retrieval-augmented generation (RAG) architecture, leveraging the strengths of large language models (LLMs) enhanced by domain-specific document embeddings. The solution was built using Python for back-end orchestration, AstraDB as the vector store, Langflow for flow logic orchestration, and a Vite-powered front end for interactive user experience. The system supports both ad-hoc and guided question-answering use cases and was implemented in a modular and extensible architecture to support future scaling.

This project focused on building a fully autonomous vending machine platform integrated with robotic replenishment and delivery systems. It’s designed for high-density urban environments and controlled-access campuses (e.g., airports, campuses, tech parks). The system combines real-time computer vision, AI-driven demand forecasting, smart inventory management, and Level 4 autonomous delivery robots capable of unsupervised, pre-defined route operations.

Managed and implemented vLLM-based inference for the client’s internal data and modified the foundational model weights using internal data for advanced diagnosis. Managed the project team, as well as being involved in the architecture hands-on.

• 15 to 60 minutes — Mean Time to Detect and Resolve (MTTDR)
• 97% to 99.5% — Device Availability

Industrial and commercial edge deployments typically grow faster than the organization’s ability to manage them. Devices are commissioned inconsistently, firmware and configuration drift accumulates over time, and connectivity is often unreliable due to NAT, segmented networks, and intermittent links.

We deliver an edge device IoT management cloud built on AWS IoT Core, using MQTT as the transport layer and Sparkplug B to standardize industrial messaging, state management, and data modeling. Devices authenticate with mutual TLS and are onboarded through controlled provisioning workflows, enabling consistent identity and policy enforcement at scale. Sparkplug’s birth/death certificates provide immediate visibility into device status and session health, while the MQTT topic structure supports clean separation of telemetry, commands, and lifecycle events.

Our approach delivers faster commissioning, predictable interoperability, and significantly improved operational reliability across heterogeneous edge fleets.

• 100% — KVKK Complıance
• 100% — GDPR Complıance

KVKK and GDPR compliance were required by the holding companies.
A multinational conglomerate required analysis and implementation of the data privacy compliance project.
They needed full compliance with both jurisdictions.

• Analysis of the current situation
• Policy and procedure definition for both GDPR and KVKK technical impact
• IT implementation of anonymization and the right to be forgotten
• Dashboard and ongoing support

Both holding companies achieved full GDPR and KVKK compliance. Customer data anonymization was implemented across group entities, the data governance model was updated with enhanced privacy procedures, and staff training was delivered.

• 20% — Increase in Efficiency Overall Over the Plans
• 50 million usd — P&L Responsibility

Management of multiple projects within a single program is needed. There were many initiatives, and banks needed one umbrella to fix all of that to be able to measure whether they are well-managed and on time.

• Created a PMO structure to manage all these projects.
• Managed the communication to internal and external stakeholders.
• Created the decks and periodicals, and dashboards for stakeholders.
• Managed the daily activities of the teams.

• Increase in the completion of the projects according to the plans.
• Increase the P&L efficiency.
• Clear communication channels to stakeholders.

Credit analysis for companies takes 4 days to prepare.
Requires manpower.
An error-prone process and a heavy load on people.caused customer dissatisfaction.
Furthermore, the regulation required the bank to respond to inquiries within the SLAs.

Design of the agentic workflow system.
On-prem implementation of the agentic workflow.
AIML ops management of the GPU infra for the project.
Governance model and the guardrails.
Tool management and integration to internal systems.

• Instead of 4 days 4 minutes analysis
• 0% error in reporting
• 100% provenance on the analysis
• 40% increase in NPS
• Compliance with standards