Skip to content
View pannet1's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report pannet1

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
pannet1/README.md

pannet1

Python · Automation · APIs · FastAPI · Integrations · AI / LLM

I turn manual workflows into services that run themselves — FastAPI backends, broker & third-party integrations, and local LLM tooling. Built for traders, operators, and anyone who wants fewer clicks.

Python FastAPI Automation APIs Integrations AI/LLM


What I ship

Capability What it looks like in practice
Python Automation Watchers, headless pipelines, schedulers and CLIs that handle files, trades and data without babysitting — open-converter (watch → ffmpeg → whisper), git-bird (audit & sync 100s of repos)
APIs & FastAPI Production FastAPI services with auth, APScheduler watchdogs, systemd + nginx in front — cookie-cutter in fastapi-apps, live in dealer-web
Integrations Broker APIs (Zerodha / Angel One / Finvasia / Delta), Telegram, WebSocket streams, ffmpeg/whisper, Google NotebookLM — see broker-ai, telegram-trades, notebooklm-py
AI / LLM Local-first agents, model chaining & memory. orchestrator runs feature scaffolding across models, rustjiin-japan keeps conversation memory with a Rust vector store on rusqlite

★ Spotlight — Desktop Buddy · Japanese Language Companion

A transparent, always-on-top desktop mascot that teaches Japanese as Itachi Sensei — offline LLM, voice in/out, and contextual memory. Built twice to explore the tradeoff:

rustjiin-japan · Rust + Tauri
Current flagship — polished & shipped

Rust · Tauri · Axum · HTMX · rusqlite vector memory · Web Speech API
Frameless transparent window, idle / thinking / listening / talking states, local LM Studio backend (Q4_K_M).

Live page → · Repo →

japanese-teacher · Go + Wails
Rebuild — Go-native, offline-first

Go · Wails v2 · templ · htmx · offline inference (Local OpenAI / CLI / Mock) · JLPT-aware prompts
Feature-sliced design (core/ + llm/), glassmorphism UI, embedded HTMX — no runtime JS bundle.

Repo →

Offline by design — no cloud keys. Plug in LM Studio on http://localhost:8080/v1 and Itachi remembers you.


Selected builds

Automation & Integrations

  • broker-ai — broker-agnostic Python lib: unified symbols + websockets for Delta / Finvasia / Angel / Flattrade (YAML-driven). pip install broker-ai.
  • open-converter — drop a video in data/ → ffmpeg → faster-whisper.txt. Idempotent, one-file-at-a-time watcher.
  • git-birdgh + git powered sync: clones missing repos, prunes archived/duplicate language folders, reports WIP / dirty / orphaned.
  • telegram-trades — Telegram calls → validated broker orders.

APIs & FastAPI Services

  • fastapi-apps — cookie-cutter FastAPI + APScheduler watchdog (market-hours scheduling, watchdog every 60s, systemd service, HTTP Basic Auth).
  • dealer-web — FastAPI dealer terminal for Indian markets (Jinja + SQLite, live order/position views).

AI / LLM Tooling

  • orchestratoragents/orch.py — model chaining, feature scaffolding (Schema/Handler/Controller/Tests), spec-QA & pytest gates. Run from any project via .agents symlink.
  • notebooklm-py — unofficial Python API + agentic skill for Google NotebookLM (capabilities beyond the web UI, CLI + Claude/Codex/OpenClaw).
  • browser-agent — Python browser automation with LLM in the loop.

Stack

Python FastAPI Go Rust Docker SQLite JS Tailwind

Python · FastAPI · APScheduler · WebSockets · HTMX · templ · Tauri · Wails · rusqlite · ffmpeg · faster-whisper · systemd · nginx · uv


GitHub Stats

stats top langs
streak


Let's build

I collaborate on automation, API integration, and AI tooling — especially where a FastAPI service or a broker/LLM integration saves hours a week.

Telegram   ecomsense.in · India

Fun fact: I sleep on my laptop.

Pinned Loading

  1. make-videos make-videos Public

    Shell

  2. open-converter open-converter Public

    open-converter

    Python

  3. orchestrator orchestrator Public

    AI coding agent orchestrator: model chaining, feature scaffolding, and agentic execution

    Python

  4. montage-is-open montage-is-open Public

    Python

  5. rustjiin-japan rustjiin-japan Public

    Rust