Independent AI studio · Morocco
Building AI
without the usual limits.
Vertex is an independent AI studio built around one idea: powerful intelligence should be useful, understandable and accessible — even when the hardware, network and budget are not unlimited.
vertex / workspace systems online
MULTIPLE AGENT SYSTEMSTRONG MEMORYMULTI-PROVIDERREAL TIME WEB SEARCHMODERN INTERFACE
01One builder.
Research & direction
One builder.
Many systems.
VertexLab is intentionally small right now: one 17-year-old Moroccan developer and a growing collection of AI agents, tools and experiments. The studio is virtual by design, with software doing much of the heavy lifting.
“
The goal isn't to make another chatbot. It's to build an AI environment that can reason across tools, remember useful context, work with files and keep evolving.
02A workspace for
Vertex AI Platform
A workspace for
real AI work.
The current platform combines conversational AI with routing, memory, knowledge ingestion, projects, web search, agents and settings. The interface stays simple while the architecture grows underneath it.
Current build
Vertex AI Platform
A modular AI workspace designed to connect users, models, tools and projects in one place.
What it actually is
Vertex AI Platform
A multi-model AI workspace with routing, memory and agents — built by one 17-year-old in Morocco.
01Multi-provider routingRequests are routed to external model providers through one orchestration layer, with automatic fallback.
02Memory + file knowledgeUseful context and project files stay attached to your account so conversations can build on each other.
03Agents and toolsPlanner, coder, reviewer and research roles, plus web search and project workspaces.
04One account, online workspaceThe same login works across Vertex AI and Vertex Characters. The app needs an internet connection — it is not an offline, on-device model.
03Designed as a
Capabilities
Designed as a
system, not a feature.
Every capability is a building block. New models, tools and interfaces can be added without turning the product into a collection of disconnected demos.
Multi-provider intelligence
Provider abstraction, model registry, task-aware selection and automatic fallback are designed into the core instead of being hard-coded into the chat UI.
Memory + RAG
Store useful context and index project files so conversations can become grounded in the user's own work.
Agent runtime
Planner, coder, reviewer, tester, research and data-focused roles provide a foundation for multi-agent workflows.
Web intelligence
Search and rich web context can be routed through the same orchestration layer instead of bypassing it.
Developer platform
Public API foundations include scoped keys, permissions, rate limits, usage accounting, workflows, webhooks, logs and documentation — designed to become the bridge between Vertex and other products.
POST /api/v1/chat
AUTH Bearer vx-••••••••
MODEL auto → routed
AUTH Bearer vx-••••••••
MODEL auto → routed
04Simple on the surface.
Architecture
Simple on the surface.
Layered underneath.
Secrets stay in the environment. Provider credentials are abstracted behind a key manager. Routing, health, metrics and response sanitization sit between the interface and external model providers.
05What comes
Roadmap
What comes
after MVP 1.0.
These are product directions, not promises. The aim is to grow the platform carefully while keeping it useful on realistic hardware and networks.
Vertex AI Platform
Chat, model routing, agents, memory, file knowledge, web search, projects, settings and developer foundations.
01Vertex Coding Agent
A dedicated coding workspace for repository understanding, changes, tests, reviews and long-running development tasks.
02Vision + Image
Multimodal understanding and image-generation workflows, with careful model and cost selection behind the scenes.
03Workflow Studio
Visual, reusable workflows that connect models, agents, tools, conditions, transformations and human approvals.
04Developer Network
Public APIs, documentation, free or accessible developer access where sustainable, and tools for builders creating on top of Vertex.
0506A studio with
The studio
A studio with
one human at its core.
VertexLab is currently a small, independent studio: Bakre Lamcharki leads product, engineering and design, while AI agents support research, coding, review, testing and other workflows.
BL
FOUNDER · BUILDER
Bakre Lamcharki
Shelby______x1 · Morocco
17-year-old developer building accessible AI systems. Leads product, engineering and design at Vertex.
DeveloperDesignerAI Builder
V
VIRTUAL TEAM
Vertex Agent System
A coordinated set of specialized AI roles that handle planning, coding, review, research and testing — supporting the builder around the clock.
PlannerCoderReviewerResearchTester
FAQ
Questions people
usually ask.
Contact Vertex
Let's build what
doesn't exist yet.
Want to collaborate, support the project, discuss developer access, investigate the build, or simply tell us what Vertex should become?
Direct contact · Morocco · Independent project