PROJECTS AND PRODUCTS

Products and enterprise solutions we have built.

Selected examples from SaaS products, custom software projects and AI-supported systems.

Glowing wireframe face model of the 3D digital twin application on a dark background.
Illuminated wireframe face model of the digital twin, with a pipeline panel on the left showing speech recognition and speech synthesis steps. Full body 3D avatar of the digital twin, with a chat panel on the right and voice and view controls below.

OUR PROJECT · ARTIFICIAL INTELLIGENCE

3D Digital Twin

A 3D digital twin that models a person or an object and talks in real time. It recognizes the speech of the user, generates an answer with a language model and speaks the answer in sync with the facial movements of the avatar. It runs in the browser without installation, the model stays on in-house hardware and the data does not leave the organization. It can be adapted to your organization for use cases from customer service to training.

Goal

To run a talking digital twin in the browser in real time, one that represents a person with their own face and voice or models an object in 3D such as a heart, a lung, a fish farm, a production line or a building. Main use cases:

  • Multilingual customer service and kiosk applications
  • Medical and vocational training
  • Museum and exhibition guides
  • Voice-based monitoring of facility and building data
  • Internal knowledge assistants
Benefit

Speech recognition, the language model and speech synthesis work in a single flow.

  • The model runs on local hardware, data stays inside the organization and GDPR and KVKK compliance is easier
  • No installation in the browser, the same experience on every device and in every country
  • Serves many users at the same time
  • Higher accessibility through voice and face interaction
  • Portrait or full body view, camera-based face tracking and a presentation mode
  • Digital Twin
  • 3D Avatar
  • Artificial Intelligence
  • Speech Recognition
  • Speech Synthesis
  • Local LLM
Connection discovery, sales opportunity details and add-customer screens of the AI-supported customer and sales management app.

OUR PROJECT · R&D · MOBILE APP

AI-Supported Customer and Sales Management

An application designed to discover potential customers, evaluate relevant business connections and track sales opportunities in a single mobile experience.

Goal
To make the process from customer discovery to opportunity tracking more organized, visible and manageable.
Benefit
AI-supported connection suggestions make research easier. It brings customer information, suggested actions, meeting history and sales documents together to help teams decide on the next step.
  • Artificial Intelligence
  • CRM
  • Sales Management
  • Customer Discovery
  • Mobile App
  • TÜBİTAK
AI Agent integration architecture diagram: web, mobile, portal and support channels on the left, an AI Agent layer in the middle made of the language model, MCP and API connections, permission policy, logging and human approval, and ERP, CRM, database, document and ticketing systems on the right.

ENTERPRISE SOLUTION · ARTIFICIAL INTELLIGENCE

AI Agents for Your Organization and Integration into Existing Software

AI Agent solutions (LLM-based agents) that take part in your business processes and work together with the software you already use. Agents connect to your ERP, CRM, issue tracking, database and documentation systems with defined permissions, pass critical steps through human approval and log every action. We do not deliver a ready-made package. We design the solution together with your systems, your data policy and your security requirements.

Purpose

Move artificial intelligence beyond the demo stage and place it safely into the daily operation of the organization. Main application areas:

  • Adding an agent layer to existing applications: the user asks in natural language, the agent acts in your systems
  • Internal knowledge assistant: documents, records and process knowledge from one point, in line with access rights
  • Request, support and document processing: classification, summarization, routing and report preparation
  • Coding agents for software teams: a plan-approved, test-driven and reviewed development routine with IDE and terminal agents
  • Team training and adoption: advanced prompt patterns, context engineering and safe usage habits
Benefit

Integration is done through open standards and APIs without changing the existing systems.

  • Connects to existing tools through MCP (Model Context Protocol) and APIs, with limits such as read-only database access defined from the start
  • Human in the loop: no action without an approved plan, no change without review
  • Least privilege, isolated runtime environment and rules against prompt injection
  • Data classification: which data goes to which model or service is part of the architecture decision, in line with KVKK, GDPR and company policies
  • Open-source or commercial model selection based on data structure, security requirements and cost, on on-premises hardware where needed
  • Measurable results through traceability, evaluation and monitoring: productivity that is proven, not just felt
Frequently Asked Questions

Can we add AI agents without changing our existing software? Yes. The agent layer is installed next to your current application and connects to your systems through MCP and APIs. Large changes in the source code are not required. Where needed, only a thin integration layer is written.

Does our data leave the company? This is an architecture decision made together with you. Models can run on on-premises hardware, classification rules define which data goes to which model or service, and confidential data is excluded from the agent from the start.

What happens if the agent performs a wrong action? Critical actions do not run without human approval. The agent works with least privilege, every action is logged and the work proceeds in reversible steps.

How do you support KVKK and GDPR compliance? The processing of personal data is designed according to KVKK, GDPR and company policies. Our team in İzmir, Türkiye runs the process from analysis to live operation and works in English and German on international projects.

  • AI Agent
  • LLM Agents
  • LLM
  • MCP
  • Enterprise Integration
  • GDPR and KVKK
  • Team Training