Skip to content
View tarunlnmiit's full-sized avatar
21 followers · 6 following

Achievements

Achievement: Starstruckx2Achievement: Pull Sharkx2Achievement: YOLOAchievement: Arctic Code Vault Contributor

Achievements

Achievement: Starstruckx2Achievement: Pull Sharkx2Achievement: YOLOAchievement: Arctic Code Vault Contributor

Block or report tarunlnmiit

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.

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
tarunlnmiit/README.md

Hi, I'm Tarun 👋

I turn scattered AI capabilities into tools people can actually run.

AI Systems Architect · LLM & MLOps Consultant — 10+ years engineering production systems, 5+ years specialised in AI/ML · GenAI · MLOps across 5 industries (Utilities · Manufacturing · E-mobility · Retail · B2B).

🇩🇪 Nearly 5 years living and working in Germany — Magdeburg · Hamburg · Düsseldorf (2019–2023, 2025–2026). MSc Digital Engineering, Otto-von-Guericke-Universität Magdeburg. EU Blue Card with fast-track re-issue, relocating back.

Portfolio LinkedIn

Production AI, measurable outcomes

  • 💰 $100K+/month saved via an LLM invoice-processing pipeline (Azure Document Intelligence → LLM → Snowflake)
  • 📞 ~65% of inbound calls auto-resolved by voice AI agents (STT → LLM → TTS, tool-calling into order management)
  • 🔁 7 production RAG chatbots and 5 MLOps forecasting pipelines — each family shipped from a single codebase
  • 🎓 IEEE-published XAI research (IRTEX) · MSc Digital Engineering (Magdeburg, Germany)

Open-source tools — installed, not just starred

Tool What it does Distribution
autopilot-jobhunt Stars PyPI Autonomous job-search agent: scans 130+ careers pages nightly, LLM-scores every role 0–100 against your full resume, drafts tailored applications. $0 API cost via 4-model fallback. PyPI · MCP (Glama) · Docker · Raspberry Pi
inbox-to-action PyPI One-pass agentic inbox triage: classify, summarize, extract tasks, draft replies — never sends (enforced by test). BYOK across 6 LLM providers incl. fully-local Ollama. PyPI · MCP server · Docker/GHCR
fable-mode Stars Behavioural discipline layer for Claude Code: session hooks inject one always-on ruleset across Opus, Sonnet and Haiku. Five recurring LLM failure modes, five enforced rules, ~117 tokens. Claude Code plugin marketplace
statusline-plus Stars Model-agnostic Claude Code status line: a single jq pass per render instead of one per segment, 3 switchable themes, merge-safe settings.json patching. Claude Code plugin marketplace
content-machine Autonomous content studio: 150 Python scripts, per-task LLM model routing, HyperFrames video pipeline, APScheduler auto-publishing — 200+ pieces/month across 7 platforms. Self-hosted
MindMorph Multi-agent adaptive learning platform: 14 specialised agents across 6 LangGraph workflows, mastery-gated progression, live in-browser code grading, RAG over uploaded PDFs. Open source

Stack

Python LangChain LangGraph CrewAI RAG MCP PyTorch AWS (Bedrock · SageMaker) Azure OpenAI Snowflake (Cortex · Snowpark) Kubernetes Terraform Docker FastAPI Kafka Airflow

Elsewhere

✍️ 252 articles on Medium — the technical ones have drawn 500K+ lifetime views, led by Data Preprocessing in Python in Towards Data Science at 148K views · ▶️ Breath of Data Science on YouTube · 🎙️ Live gesture-AI demo at WETEX Dubai

Pinned Loading

  1. autopilot-jobhunt autopilot-jobhunt Public

    AI job agent: scans 130+ careers pages nightly, scores every role against your resume with an LLM (0–100), alerts you on Telegram, and drafts tailored cover letters + resumes. Free & open source.

    Python 189 47

  2. inbox-to-action inbox-to-action Public

    One-pass agentic inbox triage: classify, summarise, extract tasks and draft replies in a single run. Draft-only — no send scope exists, enforced by test. BYOK across 6 LLM providers incl. fully-loc…

    Python

  3. MindMorph_Learning_Platform MindMorph_Learning_Platform Public

    Multi-agent adaptive learning platform: 14 LangGraph agents across 6 workflows, mastery-gated progression, live in-browser code grading, RAG over your own PDFs.

    Python

  4. content-machine content-machine Public

    Autonomous content studio: 150 Python scripts, per-task LLM model routing, video pipeline, scheduled auto-publishing — 200+ pieces/month across 7 platforms.

    HTML

  5. fable-mode fable-mode Public

    Fable-distilled behavior rules for Claude Code — Fable-class discipline on Opus, Sonnet, and Haiku

    JavaScript

  6. statusline-plus statusline-plus Public

    Model-agnostic Claude Code status line: one jq pass per render instead of one per segment, 3 switchable themes, merge-safe settings.json patching.

    Shell