Technical Assitance Required

We’re building an AI automation platform that orchestrates workflows across multiple SaaS apps using LLM routing and tool calling for JSON schema filling. Our AI stack includes:

1️⃣ Decision Layer – Predicts the flow (GET, UPDATE, CREATE) 2️⃣ Content Generator – Fetches online data when needed 3️⃣ Tool Calling – Selects services, operations & fills parameters 4️⃣ Execution Layer – Handles API calls & execution

We’re struggling with latency issues and LLM hallucinations affecting workflow reliability. Looking for fresh insights! If you have experience optimizing LLM-based automation, would love to hop on a quick 30-min call.

Please provide your thoughts ?