MistByteAI — practical local AI tooling (Search API, RAG, time plugins) • Node.js • Linux

Links

Project hubs and profiles.

GitHub (mistbyte-ai) Reddit (MistByteAI) Freelancer.com profile

For business inquiries: use Freelancer. For open-source: GitHub is the canonical source.

Open-source project: Search API V1 (WebSearcher)

A lightweight local web search API designed to feed LLMs with better context. Focused on engineering: predictable output, controllable pipelines, and clean integration.

What it does

Local web search for LLMs (search → fetch → extract → normalize)
Designed for Text Generation WebUI integration
Self-hosted Node.js service with a simple API
Emphasis on reliability, debuggability, and minimal dependencies

GitHub repository More projects

Listed in the official Text Generation WebUI extensions directory .

If this project saves you time or helps your workflow, you can optionally support its development below.

Screenshots

WebSearcher MistByte extension settings in Text Generation WebUI — Text Generation WebUI integration — settings panel

WebSearcher search result example in Text Generation WebUI chat — Example result in WebUI chat (triggered search)

WebSearcher selecting 3 links from 10 for fetching — Selection step: pick 3 links from 10

WebSearcher fetched and extracted page text preview — Fetch + extract: page text preview

Open-source project: Time Injector (CURRENT_DATE / CURRENT_TIME)

A tiny plugin that injects the current date/time into prompts so even smaller local models keep a sane timeline. Designed to be deterministic and machine-readable (no fluff, no “source”).

GitHub repository Support open-source work

Support MistByteAI open-source development

These are independent open-source projects (Search API V1 / WebSearcher, Time Injector, and more). If you find them useful, your support helps me keep improving reliability, extraction quality, and integrations.

USDC / EVM (Ethereum / BNB Chain)

0x50f859193a7f314df994b8be152661f0fa7064c8

Same address works on EVM networks (Ethereum / BNB Chain). Please double-check the network before sending.

USDC / Solana

4MMz9PvPrmRR88RvB59JhdQrzqnMkj73ByYd5DM3wxZa

Please double-check the network (Solana vs EVM) before sending.

Commercial case study (closed-source): Low-latency trading backend & real-time Telegram WebApp

Commercial case study demonstrating a custom system focused on stability, low latency and predictable transaction lifecycle control. The UI is a Telegram WebApp with real-time charts and trading controls.

Demo video

Video shows: multi-chart layout, buy/sell flow, scale updates, trailing stop & stop controls, quick toggle actions, a settings change, and live runtime logs.

What was built

Node.js backend with asynchronous state management and failure handling
Optimized transaction preparation to avoid unnecessary on-chain calls at execution time
Custom transaction lifecycle control (TTL / block reference logic)
Event-driven confirmation tracking via streaming data (gRPC) to reduce reaction latency
Real-time WebSocket transport to UI (Telegram WebApp)
1-second candle generation and charting (TradingView)

Source code is not published (internal logic / NDA constraints). Small non-sensitive snippets can be shown on request.

Screenshots

Mobile UI examples (multi-chart layout and single-chart view).

PumpBot WebApp — single chart view — Single-chart view (focus mode)

PumpBot WebApp — multi-chart layout — Multi-chart layout (4 charts)

Other experience

Joomla-based portal with deep PHP customization and data aggregation across multiple modules
Linux server infrastructure: routing/firewall, VPN/proxy setups, and network troubleshooting
Blockchain work: integrations on Solana/Ethereum; Ethereum smart contracts (custom logic and upgrades)
LLM systems integration: local search pipelines, context preparation, and WebUI extensions

Engineering background

Started programming long before modern frameworks became mainstream. Worked with a wide range of languages and platforms over the years.

Current focus: Node.js / JavaScript backend systems, LLM integration & tooling, Telegram-based solutions, PHP for legacy and custom platforms.