Not in stealth mode
This is the official public website of Mert Bileydi MBAI Automation. The company, product direction, technical roadmap and contact information are openly available.
Official company website • not in stealth mode
Mert Bileydi MBAI Automation
Applied AI lab building toward AGI through memory, reasoning, tool-use, autonomous creative operations and smart living AI.
K.U.B.R.A. AGI - Genesis V14
The AGI ambition is explicit. Our goal is to build an AGI-oriented architecture. The current stage is an early functional prototype / MVP research track focused on persistent memory, knowledge units, retrieval-augmented reasoning, self-reflection, approval-gated actions and scalable cloud infrastructure.
Company & Product Verification
Clear startup identity for reviewers, partners and cloud programs.
This section directly addresses company verification: who we are, what we are building, whether we are in stealth mode, and what online materials are available.
Own product prototypes
The company is building its own AI product tracks. It is not a cloud service provider, reseller, distributor or outsourced development agency.
Main product track
K.U.B.R.A. AGI - Genesis V14 is an AGI-oriented approval-gated AI decision architecture prototype.
Real-world demo
Lina Kaya is used as a live AI operations case study for creative campaigns, audience growth, link conversion and outreach management.
For Google for Startups Cloud review
Website: https://www.mertbileydi.com
Product: K.U.B.R.A. AGI - Genesis V14
Product type: AGI-oriented AI decision architecture prototype
Current stage: early functional prototype / MVP research track
Cloud purpose: model evaluation, high-memory inference, LoRA experiments, multimodal testing, retrieval systems, monitoring and reliable demo deployment.
Open standalone verification page
AGI Mission
We are not hiding the ambition: the goal is AGI.
K.U.B.R.A. is being built as an AGI-oriented architecture, not as a simple chatbot wrapper. The long-term objective is a system that can preserve memory, transform information into knowledge units, reason over context, use tools, reflect on its own outputs, ask for human approval when required and connect to real-world workflows.
To keep the project credible, we separate ambition from current status: the goal is AGI; the current stage is a serious early prototype and research track toward that goal.
Path toward AGI
- Persistent project memory
- Knowledge unit formation
- Retrieval-augmented reasoning
- Self-reflection and evaluation
- Tool-use and controlled autonomy
- Human approval and audit trails
- Cloud GPU scaling for larger models and demos
K.U.B.R.A. AGI - Genesis V14
Knowledge Unit Based Reasoning Automaton
A working-title AI decision architecture designed to move from raw data to structured knowledge, reasoning, reflection and controlled action.
Memory
Persistent project memory for context, continuity and long-term work.
Knowledge Units
Information is structured into retrievable units instead of being treated as disposable chat history.
Retrieval
Embeddings and vector retrieval are used to bring the right context into reasoning.
Reflection
Self-evaluation loops support safer answers, iteration and model-improvement data.
Approval Gates
Critical actions are designed to require human permission, not silent automation.
Cloud Scaling
Cloud GPU infrastructure is the next layer for heavy inference, evaluation and demos.
Evidence & Case Study
Lina Kaya is the first public AI operations layer.
Lina Kaya is not only a creative project. It is the first measurable demonstration layer for autonomous campaign operations: music rollout, ads, follower growth, Linktree conversion, DJ / beach club outreach and campaign analytics.
28.3K+Instagram ad impressions tracked
+433Follower growth measured in campaign period
42+DJ / beach club outreach targets reached
LiveMusic, social, link and outreach workflow
Lina Kaya
Autonomous AI artist and campaign manager demo for creative operations.
Living Home AI
Smart living AI track for residential automation, sensors, scenario logic and human-controlled real-world interfaces.
K.U.B.R.A. Genesis V14
The decision architecture that connects memory, reasoning, approval and execution across the lab tracks.
Infrastructure Roadmap
Scaling from prototype to reliable AI operations.
The current prototype uses a hybrid setup: a local laptop works as the control console, while GPU-heavy workloads require cloud infrastructure for high-memory inference, model evaluation, LoRA experiments, multimodal testing, retrieval experiments, monitoring and demo deployment.
Cloud infrastructure is not the product being sold. It is the scaling layer required to develop and evaluate the product properly.
Why cloud
High-memory inference, model evaluation and reliable demos need more GPU resources than the local laptop can provide.
Why now
The team is already past pure setup. The immediate need is scalable evaluation, iteration and demonstration.
Why credible
The AGI goal is explicit, while the current state is presented honestly as an early prototype and research track.
Contact
For startup program review, technical partnership or AI lab discussion.
Mert Bileydi — Founder & CEO, Mert Bileydi MBAI Automation