---

🧑‍💻 Who Am I?

class SanjidHasan:
    def __init__(self):
        self.username    = "Sanjidh090"
        self.education   = {
            "degree":      "B.Sc. in Electrical & Electronic Engineering (3rd Year)",
            "institution": "Khulna University of Engineering & Technology (KUET)",
            "country":     "Bangladesh"
        }
        self.expertise   = [
            "Deep Learning & Neural Networks",
            "Computer Vision & NLP",
            "Automatic Speech Recognition (ASR)",
            "Data Science & Analytics",
            "Embedded Systems & PCB Design"
        ]
        self.current_focus = [
            "LLM Fine-tuning", "Bengali Low-Resource NLP",
            "Generative AI", "Edge AI Systems"
        ]
        self.open_to     = [
            "Research Collaborations", "Internships",
            "Freelance / Contract Work", "Full-time (Post-Graduation)"
        ]
        self.philosophy  = "True mastery is demonstrated by the ability to teach."

    def say_hi(self):
        print("Thanks for visiting! Let's build something amazing together 🚀")

me = SanjidHasan()
me.say_hi()

---

🏆 Achievements at a Glance

---

📄 Research & Publications

🔬 arXiv Preprint — February 2026

Make It Hard to Hear, Easy to Learn: Long-Form Bengali ASR and Speaker Diarization via Extreme Augmentation and Perfect Alignment Sanjid Hasan, Risalat Labib, A H M Fuad, Bayazid Hasan

Submitted to DL Sprint 4.0 @ BUET CSE Fest 2026. Introduces Lipi-Ghor-882 — a comprehensive 882-hour multi-speaker Bengali speech dataset — and demonstrates that targeted fine-tuning with synthetic acoustic degradation outperforms raw data scaling for low-resource ASR.

Key Contributions:

• 🗃️ Lipi-Ghor-882 Dataset — 882 hours of multi-speaker Bengali audio (HuggingFace: Lipi-Ghor-bn-882-SSTT)
• 🎙️ Whisper-based ASR pipeline with extreme augmentation strategy
• 👥 Speaker diarization revealing failure modes of SOTA open-source models on Bengali data

📊 Public Datasets

Dataset	Platform	Domain
Lipi-Ghor-bn-882-SSTT	HuggingFace	Bengali ASR / Speaker Diarization
Bangladesh News Dataset	Kaggle · ResearchGate	NLP / Text Classification
KUET Whispers Dataset	Facebook Analytics	Social Media Speech
BUETIAN RAPSODY Dataset	Social Media Mining	Multimodal
Google Maps Dataset — Khulna	Geospatial	Location Intelligence

---

🎖️ Competition Record

💡 Note: AgriYield 3rd place was a solo submission — no team, no shortcuts.

---

💻 Technical Stack

---

📚 Academic Repositories

Repo	Course	Description
CSE-2132	Algorithms & Data Structures	Week-by-week DSA in C++ — Linked Lists, Stacks, Sorting, Hashing
EE-1222	Embedded Systems	Product Management & Billing System · Custom PCB design

---

✍️ Technical Writing

Article	Platform	Topic
Painting with Pixels: The Mathematics of Style Transfer	Medium	Gram Matrices, loss functions, artistic AI
What is Invariance?	DEV.to	Translation, Rotation & Scale Invariance in CNNs
Gigabytes vs. Gibibytes	DEV.to	Decimal vs. binary storage standards
The Enigma of </>	DEV.to	Coding culture & semiotics

“I don’t simply consume technology — I deconstruct it, optimize it, and teach it.”

---

⚡ Team Frozen Voltage

An active, multidisciplinary student team at KUET pushing boundaries across AI, robotics, embedded systems, and competitive data science.

Core Members: Sanjid Hasan (Founder & AI Lead) · Golam Rabby (Data Science & Math) · Shahriar Kamal (Co-representative)

What we build:

• 🤖 Robotics — Line-following robots with custom PCB design
• 🧬 Biomedical — Adaptive BAMs (Bioresorbable Acoustic Microrobots) @ SciBlitz 1.0
• 📊 Datathons — Consistent national top-10 finishes
• 📖 Technical writing — Award-winning math & literature content

---

🗺️ Google Maps — Level 7 Local Guide

100+ Followers · 1,000+ Contributions · Level 7

“I view my Maps profile as a digital archive — preserving the vibes and cultures of places that have been removed or closed. These photos become relics on maps, memories that would otherwise be lost to time.”

From quiet corners of KUET to the streets of Narsingdi and Khulna — one contribution at a time.

---

🎓 Certifications

• ✅ AI Agents Intensive Course — Google (2025)
• ✅ AIML Workshop — INDcon 2025 @ MIST
• ✅ Python — Kaggle Learn
• ✅ Intro to Machine Learning — Kaggle Learn

---

📊 GitHub Analytics

---

🎯 2026 Focus

const focus2026 = {
    research:     ["Bengali Low-Resource NLP", "Long-Form ASR", "Speaker Diarization"],
    learning:     ["Reinforcement Learning", "Vision Transformers", "Edge AI"],
    building:     ["Production-grade ASR pipelines", "Embedded AI systems", "Open datasets"],
    contributing: ["Open Source", "Technical Writing", "Community Education in Bangladesh"],
    goals:        ["Publish more research", "Top-10 Kaggle competitions", "Tech conference talks"]
};

---

🌐 Connect

---