Software Engineer specializing in backend systems, distributed architectures, and AI-powered applications. Experienced in designing microservices, building scalable APIs, and deploying cloud-native systems using Docker a…
Software Engineer specializing in backend systems, distributed architectures, and AI-powered applications. Experienced in designing microservices, building scalable APIs, and deploying cloud-native systems using Docker and Kubernetes. Strong background in Python and C++ with solid foundations in data structures, algorithms, and Linux systems. Built production-ready applications including RAG platforms, agent systems, and distributed media pipelines. Passionate about building reliable, high-performance systems and learning new technologies.
Ongoing work on conditioning mechanisms and noise schedules to improve generation fidelity and controllability., Working on data curation, 3D docking, and property-constrained training to surpass existing state-of-the-art approaches., Redesigned the CIGIN architecture by replacing Set2Set pooling with multi-head global attention and Graph Transformer layers to improve long-range molecular interaction modeling.
<p>Implement and modify deep learning modules, debug training issues, and improve documentation and usability.<br>
Collaborate with distributed developer communities through issue triage, pull requests, and code reviews.</p>
Designed image-based recommendation system using EfficientNet and FAISS, improving search accuracy by 10%., Optimized inference latency from 5s to 2s through parallel processing and model quantization techniques., Integrated AI solutions with enterprise architecture for seamless deployment serving 500+ daily active users.
<p>•Developed a full-stack web application enabling semantic search and AI-generated answers over user documents.</p><p></p><p>•Implemented hybrid search (BM25 + vector embeddings) with Reciprocal Rank Fusion for improved retrieval accuracy.</p><p></p><p>•Built backend services using FastAPI with REST APIs, asynchronous document processing, and JWT-based authentication Bcrypt hashing, and PostgreSQL-based data isolation.</p><p></p><p>•Deployed containerized services on AWS EC2 using Docker, enabling scalable document ingestion pipelines.</p><p></p>
<p>•Engineered an autonomous multi-agent system to aggregate and rank technical content using OpenAI GPT-4 and Chain-of-Thought reasoning.</p><p></p><p>•Built a Curator Agent with Pydantic structured outputs to score article relevance (0-10) against dynamic user profiles, ensuring type-safe decision logic.</p><p></p><p>•Developed a RAG-ready ingestion pipeline using Docling and Feedparser to convert unstructured video transcripts and HTML into clean Markdown for processing.</p><p></p><p>•Orchestrated daily automated workflows using a custom scheduler to manage scrape-process-email pipelines, reducing information overload.</p><p></p>
<p>•Architected a distributed microservices system for asynchronous video processing using RabbitMQ message queues and containerized worker services.</p><p></p><p>•Designed queue-based backpressure and worker scaling to prevent system overload under high job throughput.</p><p></p><p>•Implemented a gateway API service with JWT authentication, secure file uploads, and request validation.</p><p></p><p>•Built worker services to consume jobs, process media conversions, and store outputs in MongoDB GridFS.</p><p></p><p>•Deployed the platform on Docker + Kubernetes, using ConfigMaps, Secrets, and rolling updates for reliability</p><p></p>
The AI-first Recruitment Operating System that unifies and automates the entire hiring lifecycle for modern enterprises worldwide.
Dr Job FZ LLC. 2026 © All Rights Reserved
