Project Description
I’m looking to partner with an experienced agentic-AI developer who can build an end-to-end system that automates customer-related decisions. The solution should:
• Take incoming customer data (chat, email, CRM records) and decide the best support action—whether to trigger a self-service response, escalate to a human agent, or upsell an add-on.
• Generate real-time sales recommendations tailored to each customer’s profile, browsing history, and purchase patterns.
• Learn continuously from past interactions so recommendations and routing logic improve over time.
Key expectations
• Solid background in automated decision-making architectures, reinforcement learning or rule-based hybrids.
• Proven experience integrating NLP pipelines and vector databases to interpret unstructured customer text.
• Clear, well-documented code (Python preferred) along with a concise README explaining how to retrain models and adjust decision thresholds.
• Deployment guidance—Docker or similar—so I can run the engine in our existing cloud environment.
Deliverables
1. Working decision engine with reproducible training script.
2. API endpoints (REST or GraphQL) for support routing and sales recommendation calls.
3. Basic test suite plus sample request/response pairs demonstrating correct behaviour.
4. Short hand-off session walking me through configuration and future tuning.
If you’ve built comparable customer support or e-commerce AI tools, I’d love to see a short demo or repo link. Let’s create a smarter, fully agentic system that keeps customers happy while driving revenue.