AI interface / local build

Rapido keeps the conversation moving.

A focused chat surface for a hosted Qwen model, with streaming output, context and latency readouts, local conversation history, optional search, and controls that stay understandable.

Built withJavaScript · HTML · CSS · ModalInteractionSSE · Local storage · Fetch

RAPIDO / FIELD MAP02 — 03

System intention

Keep the model powerful without making the surface noisy.Conversation stays central; the useful machinery—context, speed, reasoning, and settings—remains one glance away.

JSJavaScript<>HTML/CSSModal↯SSE

Streaming UX / OpenAI compatibleLocal build

How the system reads

A chat surface with its controls visible.

The project keeps the conversation as the hero while making model behavior inspectable instead of mysterious.

Screen record / 02

Conversation, configuration, response.

Three screens show the path from a blank chat to a tuned model and a live answer.

Rapido new chat interface — 01New chatThe first screen makes context, speed, and web-search state visible without competing with the prompt.

Rapido model settings interface — 02Model settingsReasoning effort, temperature, output tokens, and context window are treated as working controls.

Rapido live response interface — 03Live response stateStreaming feedback and history make the local model feel like an instrument, not a black box.

Local project / 02

Keep reading.

The next case file moves from conversation to research: discovery, extraction, and ranked signal.

Next: Finder ↗

YAHYA / EL AMRANIBack to selected work ↗