AI-Powered Website Interaction Upgrade

I’m revamping my website to make every visitor feel as though the site is reading the room—and responding in real time. To get there, I need a developer whose English is truly native so that nuanced language, tone, and UX copy all land perfectly while the technology under the hood runs on solid Natural Language Processing, Machine-Learning algorithms, and Computer Vision.

Here’s the vision in plain terms. Users should be able to speak or type naturally and have the site understand intent, pull relevant data, and even recognise images they upload—all without friction. Think: an on-page assistant that can explain products in conversational English, suggest next steps based on browsing patterns, and identify visuals (e.g., photos of items) to match them with catalog entries.

Preferred stack is Python with libraries such as TensorFlow or PyTorch for model development, spaCy or similar for NLP, and OpenCV for any real-time visual tasks, but I’m flexible if you make a compelling case. Whatever you choose must sit behind a clean REST or GraphQL API so the current front-end (React) team can hook in quickly.

Deliverables
• Proof-of-concept models demonstrating NLP query handling, recommendation logic, and basic image recognition
• Production-ready, containerised API endpoints with clear routes and authentication hooks
• Minimal front-end demo page showing end-to-end interaction on the live site
• Straightforward documentation (setup, deployment, and usage) in crisp, idiomatic English

Acceptance criteria
• User queries return relevant results in under 800 ms for 90 % of test cases
• Image recognition confidence ≥ 85 % on provided validation set
• All endpoints pass integrated Cypress tests and basic load tests (500 rps for 1 minute without failure)

If this sounds like your wheelhouse, let’s bring intelligent, human-feeling interaction to the site.