AI SOFTWARE ENGINEERING PLATFORM

ERSELMETZ
AI CORPORATION

An AI-assisted software engineering orchestration platform designed to coordinate specialized AI agents, projects, tasks, memory, providers, testing, and future development workflows.

Core orchestration foundation in development

About

ERSELMETZ AI CORPORATION is being developed as an actual software system rather than a simulated company.

The long-term goal is to coordinate multiple specialized AI agents into a structured software engineering organization while keeping the human as the final decision-maker.

Orchestration

Coordinate projects, tasks, agents, providers, and future multi-agent workflows.

Persistent Memory

Maintain persistent agent and project context using a local database.

Human Control

Important development decisions remain under human approval and control.

Architecture

The system is being built incrementally, starting with the core orchestration infrastructure. 

                         YOU
                  CEO / PRODUCT OWNER
                           │
                           ▼
                  ┌────────────────┐
                  │  ORCHESTRATOR  │
                  └───────┬────────┘
                          │
          ┌───────────────┼───────────────┐
          ▼               ▼               ▼
      PROJECTS          TASKS           AGENTS
          │               │               │
          │               │               ▼
          │               │           PROVIDERS
          │               │               │
          └───────────────┴───────────────┤
                                          ▼
                                       OLLAMA
                                          │
                                          ▼
                                      AI RESULT

Current Status

Agents

Agent abstraction, registry, capabilities, permissions, and provider association.

Providers

Provider abstraction and registry with Ollama integration.

Tasks

Persistent tasks with status, assignment, execution results, and errors.

Projects

Persistent projects with tasks associated through project IDs.

Memory

Persistent agent memory and project-specific memory.

Audit Logs

Task creation and execution events are recorded for traceability.

Roadmap

Planned capabilities. These are not yet considered implemented.

1 Project-aware task management
2 Advanced task routing and orchestration
3 Workflow engine
4 Multi-agent collaboration
5 Controlled agent tool system
6 Developer environment integration
7 QA and security agents
8 Automated review and fix loops
9 Git-based development lifecycle

Technology

Python

Core orchestration platform.

SQLite

Persistent local application state.

Ollama

Local AI model execution.

Git / GitHub

Source control and development history.

Antigravity

Primary development environment.

Multi-Agent Architecture

Foundation for future AI collaboration.