AI Engineer & Systems Builder

I build agentic AI systems
and low-level software.

B.Tech CSE @ GCEK  ·  Open to internships in AI/ML, Backend, or Systems.

Currently seeking Internship opportunities in Backend Engineering, AI/ML Research, or Systems Programming.

Discovery

Who I Am.

Second-year CSE student at Govt. College of Engineering, Kannur. I obsess over how things work at the lowest level — from CPU pipelines to LLM inference. That curiosity drives everything I build.

I focus on agentic AI systems and backend engineering, combining systems-level thinking with modern ML tooling. I ship things, not just ideas.

2+ Years of Dev
10+ Projects Built
5+ Events & Hackathons

Current Focus

  • LangGraph & Agentic AI Pipelines
  • RAG Systems & Local LLM Inference
  • Computer Architecture (COA / GEM5)
  • FastAPI & Backend Systems
Journey

Professional Experience.

2025 – Present

Campus Co-Lead

MuLearn, GCEK

Co-leading µLearn GCEK — organizing hackathons, peer learning sessions, and technical events. Ran Build with AI · Kannur 2026 (50+ participants).

2025

AI/ML Intern

Infosys Springboard

Built CircuitGuard — a two-stage OpenCV + EfficientNet-B4 pipeline for PCB defect detection. 98% accuracy across 6 defect categories.

Knowledge

Academic Path.

2024 – 2028

B.Tech in Computer Science

Govt. College of Engineering, Kannur

CGPA: 6.9. Coursework in Data Structures, OS, DBMS, and Computer Architecture. Active in hackathons and open-source outside class.

Expertise

Technical Arsenal.

Core CS & Engineering

Python / C / Java Data Structures Computer Architecture Operating Systems DBMS & SQL Object Oriented Design

Intelligent Systems

Agentic AI Workflow Machine Learning Neural Networks NLP & LLMs Scikit-Learn Deep Learning

Tools & Ecosystem

Git / GitHub Linux (Arch/CachyOS) GEM5 Simulator Streamlit & UI Hugging Face VS Code
Portfolio

Selected Deployments.

Computer Vision & DL

CircuitGuard

Problem: High latency and error rates in manual PCB manufacturing inspections.

Solution: Developed a deep learning-based localization system to detect micro-defects.

Impact: Achieved 98% detection accuracy with 2x faster throughput than manual methods.

Python OpenCV PyTorch
Agentic AI / LangGraph

dev-logger-agent

Problem: Developers lose track of daily work — no structured git history summary exists.

Solution: A LangGraph ReAct CLI agent that reads git history and auto-generates structured dev logs.

Impact: Installable globally via pipx. Reduces daily logging overhead to zero.

Python LangGraph Groq GitPython
Computer Architecture

Pipeline Analyzer

Problem: Difficulties in visualizing and mitigating data/control hazards in CPU pipelines.

Solution: A Python-based integration with GEM5 to simulate and analyze structural hazards.

Impact: Identified 3 critical bottlenecks in a dual-issue CPU simulation environment.

Python GEM5 Architecture
Milestones

Awards & Recognition.

National Finalist

HackWise 2026: Secured a top-tier finish among 500+ participants nationwide at IIM Indore.

Innovate Hackathon

IEEE Kerala 2025: Spearheaded the development of a production-style AI solution for the challenge.

Tinkerhub Sprint

18Hrs Prototyping: Rapid-prototyped a technical solution during an intensive offline sprint.

100 Days of AI

Learning Trajectory: Publicly documented a rigorous 100-day journey from fundamentals to agentic AI.

Open Source & Activity

Continuously building in public and contributing to core systems and AI repositories.

Visit GitHub Profile
Karthik's GitHub Stats
Contact

Get in touch.

Direct Uplink

I'm currently looking for internships and collaborative projects in AI and Systems.