2nd Year B.E. CSE @ GLEC | 9.05 CGPA (9.25 SGPA in Sem 3)
Hi there, I'm Varnika Mishra! 👋
I am an AI researcher and developer specializing in bridging the gap between high-fidelity hardware telemetry and cognitive reasoning engines. My work, which spans Physics-Informed ML, Multimodal AI Sovereignty, and Edge IoT, has been presented at Rashtrapati Nilayam for National Science Day 2026 and validated by T-Works.
- National Recognition: Selected to present "EV Range Intelligence" at Rashtrapati Nilayam (President's Nilayam) for National Science Day 2026.
- Academic Excellence: Consistent upward trajectory (8.86 ➔ 9.02 ➔ 9.25 SGPA) maintaining a 9.05 cumulative CGPA.
- Global Leadership: Lead Organizer of an International Technical Conference; managed cross-functional teams and international keynote speakers.
- Industry Validation: IIC Regional Finalist for "Structural Logic Innovation"; work showcased to industry experts at T-Works.
- NPTEL Honors: Gold Elite in Technical Communication (Top 1% candidate) and 2 Silver Elites in C and Python.
- Roles: Core Member at GDG GLEC; Research Lead at Advanced Academic Center (AAC) GRIET.
- Core AI & Multimodal: Gemini 3 (Flash/Pro), Long-Context Reasoning (1M+ tokens), YOLOv8, OpenCV, RAG Architecture, Prompt Engineering.
- Languages: Java, Python, & C++ (HackerRank Gold), C, SQL, HTML5/CSS3.
-
Engineering & Math: Physics-Informed ML (
$P \propto v^3$ ), Graph Theory (Nodes/Edges), Time-Series Telemetry, IoT Sensor Fusion. - DevStack: Streamlit, Plotly Express, Arduino, Git/GitHub, Cloudinary API, SDLC Workflows.
- Innovation: Achieved 94.2% accuracy by merging a Random Forest Regressor with a custom Physics Engine accounting for aerodynamic drag and thermal drift.
- Empirical Rigor: Mapped the "Lithium-Ion knee" via 50+ hardware discharge cycles.
- Cognitive Layer: Integrated Gemini 3 for sub-second, context-aware natural language trip consulting.
- Concept: A hybrid vision pipeline (YOLOv8 + Gemini 3 + EasyOCR) that eliminates visual guesswork in flowchart-to-code translation.
- Architecture: Compiles visual logic into a formal JSON Graph (Nodes/Edges), ensuring structural certainty across Python, Java, C++, and C.
- AI Proofreader: Uses Gemini 3 to autonomously correct OCR typos and interpret complex mathematical symbols.
- Architecture: A decentralized IoT-Vision Fusion model for massive scalability, utilizing IP cameras to replace expensive RFID infrastructure.
- Security: Features Gesture-Based Emergency Detection and Unusual Activity Monitoring via OpenCV for real-time security alerts.
- Process: Automated 3-step workflow (Trigger ➔ ANPR ➔ Gate Control).
- Vision-to-Sheet: Uses Pixel-Sharing methods for spatial layout analysis to digitize complex legal/financial documents.
- Vantage Juris: Leveraged Gemini 3’s 1M+ context window to architect trial strategies from landmark Supreme Court precedents.
- Accessibility: Voice-first design supporting Hindi & Marathi.
To know more about me and my works:
LinkedIn | GitHub | HackerRank
"Building the future of Bharat through secure, multimodal intelligence."