An intelligent multi-agent platform for analyzing software project requirements, generating technical specifications, and delivering project strategies β powered by OpenAI agents working in synergy.
AI Services Agency is your all-in-one virtual consultancy powered by AI. It brings together a team of specialized agents β CEO, CTO, Product Manager, Developer, and Client Manager β who collaborate to evaluate your project idea and deliver:
β
Project Analysis
β
Technical Specifications
β
Product Roadmap
β
Development Plan
β
Go-to-Market Strategy
| Agent | Role | Responsibilities |
|---|---|---|
| π§βπΌ Project Director (CEO) | Strategic Evaluation | Analyzes feasibility, goals, and alignment |
| π οΈ Technical Architect (CTO) | System Design | Defines architecture and technologies |
| π Product Manager | Roadmap Planning | Develops roadmap and product vision |
| π¨βπ» Lead Developer | Implementation | Suggests tech stack and implementation plan |
| πΌ Client Success Manager | GTM Strategy | Designs customer acquisition and delivery flow |
- π§ Multi-agent collaboration via Agency Swarm
- π Project intake form with budget, timeline, and requirements
- π Automated analysis of project feasibility
- ποΈ Generation of technical architecture specs
- π§ Product strategy insights and developer planning
- π Go-to-market plan for client delivery success
- π OpenAI API integration with token security
- π₯οΈ Built with Streamlit for rapid deployment
π¦ ai-services-agency/
βββ π agents/ # π€ Definitions for each specialized AI agent
β βββ ceo.py # π§βπΌ CEO β Strategic project analysis
β βββ cto.py # π οΈ CTO β Technical architecture and specs
β βββ developer.py # π¨βπ» Lead Developer β Implementation planning
β βββ product_manager.py # π Product Manager β Roadmap and features
β βββ client_manager.py # πΌ Client Success Manager β GTM strategy
β
βββ π tools/ # π§° Tools used by agents
β βββ analyze_project.py # π Analyze project requirements
β βββ create_specification.py # π Create technical specification
β
βββ π core/ # π§ Core app logic and shared session
β βββ session.py # π§ Session state initialization
β
βββ π app.py # π Main Streamlit app β UI + Agent Orchestration
βββ π requirements.txt # π dependencies
βββ π LICENSE # π MIT License.
βββ π README.md # π Project documentationgit clone https://github.com/andredisa/AI_services_agency.git
cd AI-services-agencypython -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activatepip install -r requirements.txtYouβll be prompted to enter it in the sidebar the first time you run the app.Start by completing a simple form with essential information:
π Name Β· π§© Description Β· π οΈ Type
ποΈ Timeline Β· π° Budget Β· π§ Technical Needs
Once submitted, a squad of expert AI agents gets to work behind the scenes:
-
π§βπΌ CEO (Project Director)
β€ Provides high-level strategic analysis and project feasibility -
π οΈ CTO (Technical Architect)
β€ Defines architecture and selects core technologies -
π Product Manager
β€ Suggests product roadmap, features, and delivery strategy -
π¨βπ» Lead Developer
β€ Plans the technical implementation and development flow -
πΌ Client Success Manager
β€ Proposes a go-to-market and customer success plan
Everything is auto-organized into interactive tabs:
π View Strategic Analysis
π§± Read the Technical Specification
π Explore the Product Roadmap
π Plan Your GTM Strategy
All in one clean, collaborative workspace.
- π OpenAI GPT Agents β for intelligent decision-making
- π§ Agency Swarm β multi-agent orchestration framework
- πΊ Streamlit β to power the sleek and interactive frontend
- β Requires a valid OpenAI API key with access to GPT-4 or higher
- π¦ Agent memory and shared state ensure consistent and sequential logic
- π§ͺ This is a prototype β simple, extensible, and ready for your creative ideas!
Here are some cool things weβd love to add (but arenβt built in... yet π):
- π§ Persistent memory across sessions
- π Exportable PDF reports of agent output
- π§βπ¨ Custom personas for specific industries
- π₯ Team collaboration & commenting dashboard
π Contributions are more than welcome!
If you find a bug π, have a feature request β¨, or want to improve the code π»:
- Open an Issue
- Submit a Pull Request π
Letβs build this together!
π This project is released under the MIT License.
Please refer to the LICENSE file for full details.
If you find my work useful and would like to support me, you can buy me a coffee! Your support helps me keep creating and improving my projects. Thank you! π