ResearchDeployed
PRISM
Offline conversational AI on Raspberry Pi with contextual reasoning and hybrid retrieval.
HI, I'M
Karthik M
Computer Science Engineer | Builder | Learner
Exploring the intersection of AI, FinTech and Real-world Impact.
“Build today
for a better tomorrow.”
About Me
More about me
I'm Karthik, a Computer Science and Engineering student at Vidyavardhaka College of Engineering, Mysuru. I'm passionate about building technology that solves real-world problems, with a focus on AI/ML, FinTech, and impactful software products.
I love exploring how intelligent systems can work on resource-constrained devices, and I enjoy turning ideas into working prototypes — whether it's for research, hackathons, or open-source.
My goal is to combine technology, research, and global opportunities to create meaningful impact and learn continuously.
Mysuru, IndiaIST (GMT+5:30)
5th SemesterCSE, VVCE
Open toInternships / Research
Currently LearningJava Backend Systems
1,790CONTRIBUTIONS
▣ LAST YEARNovDecJanFebMarAprMayJunJulAugSepOct
Preview data · connect @leafyte to show real activityLess More
ASK THE PORTFOLIO
HackathonMVP
MANAKSETU
AI-powered platform for Legal Metrology compliance of packaged commodities.
PersonalIn Progress
Sports Predictor
Machine learning model to predict outcomes for IPL, NBA and Football matches.
Research Intern (College Project)
PresentVVCE, Mysuru | Jan 2026 – Present
- Working on offline conversational AI (PRISM) using TinyLlama on Raspberry Pi.
- Designed hybrid retrieval and contextual reasoning pipeline.
- Achieved 0.884 multi-turn contextual accuracy (22.6% ↑ over baseline).
- Technologies: Python, Ollama, TinyLlama, SQLite, Raspberry Pi.
Vidyavardhaka College of Engineering
Mysuru, Karnataka
B.E. in Computer Science and Engineering
Aug 2024 – May 2028
CGPA: (Optional)
Relevant Coursework: DSA, DBMS, OS, CN, Software Engineering, AI/ML, Cryptography
JavaPythonCJavaScriptTypeScriptReactNode.jsSpring BootFlaskMachine LearningComputer VisionRaspberry PiDockerAWSSupabaseMySQLMongoDBGit & GitHub
My Resume and Portfolio Story: How I Stopped Having Two Documents That Said Different Things
How aligning my resume, GitHub, and portfolio into a unified narrative turned a scattered project list into a compelling engineering story.
Stop Typing Vague Prompts: How Students Can Get Better at Vibe Coding
Moving past slot-machine prompting — practical lessons on crafting detailed design prompts, idea validation, and building real products with AI.
How We Went From “I Think RL Is Cool” to a Research Paper (And How You Can Too)
The honest, step-by-step journey of how undergrads with zero prior research background built an RL NPC framework and published an IEEE paper.
National Finalist
Finalist / Top Innovation - Smart India Hackathon
Ministry of Education, Govt. of India
Developed MANAKSETU: Automated Legal Metrology compliance verification system with OCR and NLP.
Peer-Reviewed Paper
Research Paper Accepted at ICIDTSM 2026
IIT Madras
Published work on hybrid contextual reasoning architecture for offline conversational AI on edge devices.
Let's build something impactful together.
I am open to internship opportunities, research collaborations, and exciting projects in AI/ML, Edge Computing, and Full-Stack Engineering.