AI Agent Development for Meta

I’m planning to roll out an AI-powered agent that plugs seamlessly into the Meta ecosystem and automates day-to-day interactions on the platform. The core objective is to have a flexible framework that can be steered toward customer support, content moderation, data analysis—or a smart blend of all three—once we settle on the most valuable starting point together.

Here’s what I need from you:
• A clear technical proposal outlining the model architecture (rule-based, ML, or deep-learning driven), the libraries you recommend—think Python, PyTorch or TensorFlow—and how the agent will hook into Meta’s Graph API or other relevant endpoints.
• Production-ready code with comments and an environment file so I can reproduce your results locally.
• A lightweight front-end or command-line interface to showcase key flows.
• A short video or written demo that proves the agent is live, responding accurately, and meeting agreed-upon KPIs such as latency, precision, and scalability.

Acceptance criteria:
1. The agent must authenticate securely with the chosen Meta service.
2. Responses should remain under 300 ms for the benchmark test we’ll run on a mid-tier VM.
3. Accuracy should exceed 90 % on a mutually defined validation set.

If you’ve built conversational bots, moderation pipelines, or data insight tools for social platforms before, let me know—code samples or live links are a plus. I’m ready to get started as soon as we agree on the best approach and timeline.