Skip to content
View nibir-ai's full-sized avatar
๐Ÿ 
Working from home
๐Ÿ 
Working from home

Block or report nibir-ai

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 supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this userโ€™s behavior. Learn more about reporting abuse.

Report abuse
nibir-ai/README.md

class NibirBiswas:
    def __init__(self):
        self.name        = "Nibir Biswas"
        self.role        = "AI / ML Engineer"
        self.location    = "India ๐Ÿ‡ฎ๐Ÿ‡ณ"
        self.os          = "Linux ๐Ÿง"  # non-negotiable
        self.learning    = ["PyTorch", "Deep Learning", "MLOps", "LLMs"]
        self.passion     = ["Computer Vision", "Neural Nets from Scratch", "Open Source"]
        self.fun_fact    = "I debug neural nets faster than I debug my sleep schedule ๐Ÿ˜…"

    def currently(self):
        return {
            "building"  : "ML tools that actually solve real problems",
            "studying"  : "The math behind every layer โ€” no black boxes",
            "mantra"    : "Understand it. Build it. Break it. Rebuild it better.",
        }

    def __str__(self):
        return "Turning data into decisions, one epoch at a time. ๐Ÿš€"

me = NibirBiswas()
print(me)



๐Ÿ› ๏ธ ย  Tech Arsenal

Languages & Core

AI / ML / Data

Tools & Environment



๐Ÿš€ ย  Featured Projects

๐Ÿ‘๏ธ DatasetVision

Python OpenCV CLI

Industry-grade dataset governance CLI for computer vision pipelines. Detects anomalies, near-duplicates, data drift, corruption, and class imbalance โ€” before your model ever sees a single batch.

drift-detection cv2 data-quality typer

View Repo

๐Ÿ”ฌ micro-tensor

Python From Scratch Autograd

A tensor engine built from the ground up. Because the best way to understand deep learning isn't to use a framework โ€” it's to build one yourself. Autograd, forward/backward passes, all of it.

autograd backpropagation tensors from-scratch

View Repo

๐Ÿง  XBMind

Python AI Experimental

An intelligent reasoning system exploring the architecture of machine cognition. Where neural computation meets structured thinking โ€” an experiment in what AI can become.

reasoning agents ai-systems experimental

View Repo



๐Ÿ“Š ย  GitHub Stats








๐Ÿ“ˆ ย  Contribution Graph



๐ŸŒฑ ย  Currently On My Learning Arc

๐Ÿง   Deep Learning Theory    โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  60%   โ€” diving into the math  
๐Ÿ”ฅ  PyTorch                 โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  35%   โ€” hands-on every day  
๐Ÿ—๏ธ  MLOps & Pipelines       โ–ˆโ–ˆโ–ˆโ–ˆโ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  20%   โ€” infra for real models  
๐Ÿค—  HuggingFace / LLMs      โ–ˆโ–ˆโ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  10%   โ€” on the horizon  
๐Ÿณ  Docker / Containers      โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘โ–‘  55%   โ€” getting comfortable  


โšก ย  Fun Facts

๐Ÿง Linux is my home โ€” I don't just use it, I live in it.

๐Ÿ”ฌ I build frameworks before I use them โ€” micro-tensor exists because understanding beats importing.

๐ŸŒ™ Best debugging happens at midnight โ€” when the logs finally start making sense.

๐Ÿ“– I read papers for fun โ€” arxiv.org is my social media.

โ˜• Fuel: Coffee โ†’ Code โ†’ Train โ†’ Evaluate โ†’ Repeat.




"The best model is the one you understand completely."

โœ‰๏ธ nibir.github@gmail.com ย |ย  ๐Ÿ™ github.com/nibir-ai

Pinned Loading

  1. XBMind XBMind Public

    Turn any Bluetooth speaker into an offline AI smart speaker โ€” no hardware mods, pure Python. Wake word, Whisper STT, local LLM via Ollama, Piper TTS. Runs on Linux. Zero cloud required.

    Python 2

  2. datasetvision datasetvision Public

    Offline computer vision dataset auditing CLI tool for validating image datasets before model training.

    Python 1

  3. arguewell arguewell Public

    TypeScript

  4. micro-tensor micro-tensor Public

    A lightweight, from-scratch automatic differentiation engine.

    Python