Milos Velimirovic
Available for work

Hi, I'm
Milos
Velimirovic

|

I craft intelligent software solutions at the intersection of AI and modern web — turning complex problems into elegant, scalable products.

Scroll
ai_agent.py
import anthropic
from typing import Iterator

class AIAgent:
    def __init__(self):
        self.client = anthropic.Anthropic()
        self.model = "claude-sonnet-4-6"

    def think(self, prompt: str) -> Iterator:
        with self.client.messages.stream(
            model=self.model,
            max_tokens=1024,
            messages=[{
                "role": "user",
                "content": prompt
            }]
        ) as stream:
            yield from stream.text_stream

agent = AIAgent()
# Building the future, one token at a time
Milos Velimirovic

Turning ideas into
intelligent software

I'm an AI Software Developer with a passion for building systems that think, learn, and adapt. My work spans machine learning pipelines, LLM integrations, and full-stack web applications that bring AI capabilities to real-world users.

When I'm not training models or architecting APIs, I'm exploring the latest research in reinforcement learning and multimodal AI — always looking for the next breakthrough to put into production.

5+Years Experience
40+Projects Shipped
15+Happy Clients

AI Engineering

LLM fine-tuning, RAG pipelines, agent frameworks, and production ML deployment.

🌐

Full-Stack Dev

End-to-end web apps from React/Next.js frontends to Python/Node.js backends.

☁️

Cloud & DevOps

Scalable infrastructure on AWS/GCP with Docker, Kubernetes, and CI/CD pipelines.

🔬

Research

Staying at the frontier of NLP, computer vision, and multimodal AI research.

Tech Stack

AI & ML

PyTorch TensorFlow LangChain OpenAI Anthropic Hugging Face scikit-learn RAG Fine-tuning Vector DBs

Languages

Python TypeScript JavaScript Go SQL

Web & APIs

Next.js React FastAPI Node.js GraphQL REST

Cloud & Infra

AWS GCP Docker Kubernetes Terraform Vercel PostgreSQL Redis

Selected Work

RAG Vector DB Next.js

Intelligent Document Q&A

RAG-powered knowledge base with semantic search across thousands of documents, streaming responses, and citation tracking. Reduced customer support load by 60%.

Computer Vision PyTorch React

Real-Time Vision Pipeline

Object detection and classification system processing 30fps video streams. Custom YOLO fine-tuning, optimized inference with ONNX, deployed on edge hardware.

NLP Fine-tuning AWS

Sentiment & Intent Classifier

Fine-tuned transformer model for customer feedback analysis across 12 languages. Achieves 94% accuracy, integrated into a real-time analytics dashboard with trend detection.

Let's build something
remarkable

Whether you have a project in mind, need AI consulting, or just want to talk tech — I'm always open for a conversation.