Skip to content
View Rohit909-creator's full-sized avatar
😌
Ellaam Ariyano
😌
Ellaam Ariyano
19 followers · 17 following

Achievements

Achievement: Quickdraw

Achievements

Achievement: Quickdraw

Block or report Rohit909-creator

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.

Please don't include any personal information such as legal names or email addresses. Maximum 250 characters, markdown supported. This note will be visible to only you.
Report abuse

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

Report abuse
Rohit909-creator/README.md

Hi there 👋, I'm Rohit

Creator of A.I.R.A (An Intelligent Robot Assistant)

🚀 About Me

  • 👨‍💻 AI Engineer | Neural Architect | Robotics Innovator
  • 🧠 I design new neural network architectures, create AI tools, and build hacking gadgets
  • 📚 Btech Degree Post Graduate on Computer Science (AI & ML) from Sree Chitra Thirunal College of Engineering 2025
  • 🧪 Self-taught in Machine Learning & Deep Learning through YouTube and hands-on projects
  • 🤖 Passionate about Robotic Manipulation, MultiModal AI, SLMs, Computer Vision, Speech Recognition, LLM Research
  • 🧬 Goal: Build the first conscious AGI — I won't stop until I do.
  • Checkout my First Research Paper: https://ieeexplore.ieee.org/document/10730290

🔧 Tech Stack

📂 Featured Projects

Project Description Tech Stack Link
A.I.R.A A half humanoid robot powere using gpt4, which can talk, move, and interact with users. All of this is controlled by GPT4 because of our good prompting. AI Prompting, Python, OpenCV, PyTorch, Raspberry Pi, NLP, AI function call GitHub Repo
Speaksense Replacement for the good old wakeword detection system using a Multimodal system for detecting when someone is addressing a robot (video + audio fusion) Go to stable branch for usage ASD, NLP, CV, BiLSTM, Lipreading GitHub Repo
Human_in_the_Loop_Reasoning a finetuned DeepSeekR1 Model which asks user doubts to ambigous questions, and proceeds with its further reasoning steps. Just like token in reasoning models, here , tokens are used to ask user doubts. This can be used in any reasoning model to improve its accuracy as it reduces the ambiguity in the task assigned to the model and also I think this might reduce the compounding error in reasoning models. DeepSeek R1, HuggingFace, Unsloth, New Reasoning Way View Project
Hierarchical_Reasoning_Model Not just a simple paper re-implementation, but a vision extension and generalization comparison of HRM model with Resnet and VGG on few sample of data Wakeword Detection, Deep Learning, Audio, Real-time, Noise Adaptive View Project
SpecialDecoder a replacement layer to M to N blocks in an LLM, trying to make LLM compressed Research,LLM, Python GitHub Repo
Reward-Augmented-Decoding-RAD- Using Reward Signals to prevent Toxic Responses in LLMs Research,LLM, Python GitHub Repo
Horizon-ai-assistant GPT_OSS and Gemini+Omniparser powered multiagent framework fo automating system controls, Practical Jarvis Research,LLM, Python GitHub Repo
MrRobot An AI based Reverse Access Trojan, which can be a nightmare for some people... Hacking tool, Open-Interpretor, Python GitHub Repo

📫 How to Reach Me

🧠 Quote I Live By

“The people who are crazy enough to think they can change the world are the ones who do.” – Steve Jobs

📊 GitHub Stats

stats langs

🧪 Fun Fact

I'm on a mission to create the world's first conscious AGI.
Not just dreaming it — I'm building it. 💥

Pinned Loading

  1. Dawn-Of-Justice/SpeakSense Dawn-Of-Justice/SpeakSense Public

    SpeakSense is a multimodal deep learning project that detects when a user is speaking to a virtual assistant by analyzing both audio and video in real time

    JavaScript 2

  2. Reward-Augmented-Decoding-RAD- Reward-Augmented-Decoding-RAD- Public

    Using Reward Signals to prevent Toxic Responses in LLMs

    Jupyter Notebook

  3. EurekaBotics/project-AIRA EurekaBotics/project-AIRA Public

    This is AIRAv4 at its initial stage, It has Wave, Hi, Salute and is powered by llama3, gemini and groq

    C++ 3

  4. Hierarchical_Reasoning_Model Hierarchical_Reasoning_Model Public

    Just testing if HRM Generalizes better compared to others in Vision Task

    Jupyter Notebook

  5. Dawn-Of-Justice/Horizon-ai-assistant Dawn-Of-Justice/Horizon-ai-assistant Public

    An AI assistant that can do pretty much anything on the computer by leveraging open source LLMs

    C# 1

  6. Human_in_the_Loop_Reasoning Human_in_the_Loop_Reasoning Public

    Its a finetuned DeepSeekR1 Model which asks user doubts to ambigous questions, and proceeds with its further reasoning steps

    Jupyter Notebook 1