About Dragan
I’m a senior full-stack and AI software engineer with 8+ years of experience building scalable web applications, SaaS platforms, APIs, dashboards, and AI-powered products.
I help startups and companies design, build, and launch production-ready software using React, Next.js, TypeScript, Node.js, Django, FastAPI, PostgreSQL, AWS, OpenAI, LangChain, and vector databases.
My work includes LLM applications, RAG systems, semantic search, AI chatbots, voice agents, automation workflows, backend APIs, admin dashboards, and cloud-based platforms. I focus on clean architecture, maintainable code, reliable delivery, and practical business value.
I can help you build an MVP, improve an existing product, integrate AI features, automate internal workflows, or scale your backend and frontend systems. My goal is to turn business requirements into software that is easy to use, stable in production, and valuable for real users.
English
Fluent
Remote only
Primarily works remotely
Experience
- Xpress HealthSenior AI Solutions/Full Stack EngineerJuly 2024 - December 2025 (1 year and 5 months)Dublin, IrelandXpress Health is a• • Designed and implemented Python backend services (FastAPI) for clinical documentation and analytics AI agents, integrating Retrieval-Augmented Generation (RAG) pipelines which reduced clinician documentation time by 30% and improved provider retention.• • Built end-to-end RAG pipelines by chunking, embedding, and indexing proprietary clinical guidelines and medical policies into a vector database (Pinecone), reducing hallucinations and enabling safe enterprise healthcare adoption.• • Implemented embedding strategies and semantic retrieval logic to ensure AI responses were grounded in patient-safe, institution-approved medical knowledge.• • Built analytics AI agents that summarized patient encounters and operational data, supporting faster clinical decision making and reducing manual review effort.• • Devised and orchestrated a multi-agent architecture leveraging LLMs (GPT-4o, GPT-4o-mini, Cohere), improving intent detection accuracy by 35% and reducing misclassification in conversational flows by 40%.• • Developed evaluation workflows to measure retrieval accuracy, response grounding, and clinical relevance, enabling continuous improvement of AI outputs in production.• • Enhanced model quality through LoRA fine-tuning, evaluation workflows, and prompt optimization, increasing response accuracy and reducing latency in production.• • Built clinician-facing Next.js interfaces for reviewing, editing, and approving AI-generated clinical notes, increasing trust and adoption of AI features.• • Deployed and operated AI services on AWS (ECS, Lambda, S3, RDS), supporting concurrent clinical sessions while meeting strict latency and availability requirements.
- Vector MLAnalyticsAI/ML EngineerAugust 2023 - June 2024 (10 months)New York, NY, USAFinancial Forecasting Platform for Banks & Lending Institutions• • Built LLM-powered financial analytics agents enabling finance teams to query data in natural language, reducing time toinsight from hours to minutes.• • Developed decision-support AI systems combining GPT-3.5 with structured financial data and RAG pipelines to support forecasting and risk analysis.• • Developed semantic document search services over contracts, financial policies, and historical reports using vector similarity search, improving information retrieval speed during audits and compliance reviews.• • Designed prompt templates, JSON output schemas, and post-processing validation logic to enforce numerical accuracy, deterministic responses, and source traceability in LLM outputs.• • Implemented a voice-enabled interface using speech-to-text pipelines integrated with GPT-3.5 APIs, allowing internal finance teams to query financial data via voice and receive structured, auditable responses.• • Refined backend pipelines with FastAPI and Python, reducing response latency by 45% and enabling real-time text-to-speech with 99.9% uptime.• • Constructed and fine-tuned intelligent classification models supporting recommendation systems, wardrobe organization, and event detection.• • Rolled out GPU-based inference pipelines using Docker and cloud compute, streamlining scalability and improving system throughput.
- ClockworkSoftware Engineer (AI/LLM Systems/RAG)January 2022 - June 2023 (1 year and 5 months)Boston, MA, USAAn AI-native SaaS platform that replaces spreadsheets with an intelligent FP&A assistant delivering instant financial insights.• •Architected a production-grade semantic chatbot using LangChain and Pinecone, integrating domain NLP for real-time accuracy.• • Integrated AI-assisted search and RAG workflows using LLMs, embeddings, and vector databases to enhance functionality and provide contextual knowledge retrieval.• • Built and maintained scalable backend APIs using NestJS and FastAPI, ensuring secure data ingestion, transformation, and reliable storage pipelines.• •Architected resilient anti-blocking strategies (IP rotation, user-agent spoofing, adaptive retries), achieving 99.9% reliable production ingestion.• • Developed React/Next.js + TypeScript responsive dashboards and chat interfaces, evolving the MVP into a production platform for 1,000+ enterprise users.• • Drove significant latency and database load reductions by architecting and deploying Redis caching across high-traffic services.• • Built CI/CD pipelines using GitLab and Docker, improving deployment speed and minimizing service downtime.
Recommendations
Be the first to recommend Dragan
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Bachelor of Software EngineeringSingidunum University2017Bachelor of Software Engineering