I'm a Computer Science and Engineering student at IIT Palakkad, building strong foundations in AI and Backend Engineering through practical, well-engineered projects. My work spans optimization algorithms, distributed systems, computer vision pipelines, and full-stack AI applications.

I care about clean architecture, measurable results, and engineering decisions I can explain.

name: Ajaykumar Mallameeda
role: CSE Student
institution: Indian Institute of Technology Palakkad
interests:
  - Artificial Intelligence
  - Backend Engineering
  - Distributed Systems
  - Agentic AI
currently_building:
  - Liner Shipping Optimizer (GA + MILP, FastAPI, K8s)
  - AI Communication Assistant (Next.js, TypeScript, LLMs)
learning:
  - Agentic AI architectures
  - Distributed systems design
  - Advanced backend patterns
open_to: Software Engineering / AI Engineering roles

• Agentic AI — Multi-agent orchestration, LLM workflows, autonomous decision systems
• Backend Engineering — Scalable APIs, distributed systems, cloud-native architecture
• Optimization & OR — Metaheuristics, MILP, combinatorial optimization, vehicle routing
• Computer Vision — Deep learning for medical imaging, CNNs, object detection
• Cloud Infrastructure — Containerization (Docker, K8s), CI/CD, cloud deployment
• LLM Applications — RAG systems, prompt engineering, AI-powered tools

Distributed GA + MILP system for maritime liner network optimization

Problem: Global liner shipping networks involve complex routing decisions across multiple ports, vessel types, and cargo demands. Traditional optimization methods don't scale to real-world problem sizes.

Approach: Hybrid architecture combining a genetic algorithm for global search with mixed-integer linear programming for local refinement, orchestrated across multiple agent-based regional solvers.

Engineering challenges:

• Decomposing a combinatorial problem across distributed agents
• Coordinating parallel GA runs with MILP-based constraint resolution
• Building a reactive dashboard for real-time optimization monitoring

Tech: Python · FastAPI · Kubernetes · Docker · React · PostgreSQL

View Repository →

Note: Exploration of distributed optimization architectures. Simulation results reflect controlled experimental environments.

Adaptive signal control and vehicle routing via time-varying network flow optimization

Problem: Urban traffic congestion at intersections is managed by static timing schedules that don't adapt to real-time conditions.

Approach: A Unity-based traffic simulation with an external optimizer for dynamic signal timing. File-based IPC connects a Python optimization engine to a 3D Unity simulation environment.

My role: Simulation architecture and Unity implementation. Team project with 4 contributors across simulation, signal control, path routing, and vehicle pooling modules.

Engineering highlights:

• Multi-language, multi-process architecture (Unity + Python orchestrator)
• External optimizer generates adaptive signal timings based on traffic density

Tech: C# · Unity · Python · OpenCV · TensorFlow · YOLO

View Repository →

⚠ Known limitation: Single-intersection simulation. File-polling IPC (not event-driven). Trade-offs documented in the repository.

Deep learning grading pipeline for retinal fundus images

Problem: Manual grading of retinal fundus images for diabetic retinopathy is time-consuming and subjective. Automated screening can assist clinicians.

Approach: Two parallel pipelines — a classical image processing approach (SVM, KNN with hand-crafted features) and a deep learning pipeline using MobileNetV2 with transfer learning.

Key work:

• Feature extraction pipeline: adaptive histogram equalization, discrete wavelet transform, Gabor filters, K-means segmentation
• Transfer learning with MobileNetV2 for severity grading
• SVM: ~96% accuracy on DIARETDB1 (89-image dataset — preliminary results)

Tech: Python · PyTorch · TensorFlow · OpenCV · Scikit-learn

View Repository →

Results context: Single-dataset experiments. Not validated for clinical use. Educational and research purposes only.

Building strong foundations in AI and Backend Engineering through practical, well-engineered projects.
_{IIT Palakkad · CSE 2023–2027}