Description

I’m Victor Cecoltan, a Senior AI/ML Engineer with over 8 years of experience delivering production-grade AI systems. I specialize in generative AI, large language models, NLP, Agentic RAG systems, recommendation engines, and computer vision. What sets me apart is my ability to architect end-to-end solutions - from data pipelines and model training to scalable deployments - ensuring real-time performance and measurable business impact.

Industry field of expertise

Languages

English
Fluent
Russian
Fluent

Workplace preferences

Remote only

Primarily works remotely

Ai
Senior AI Engineer
June 2025 - Today (1 year)
San Francisco, CA, USA
• Participated in the development of an automated multimodal content intelligence system, utilizing LLMs to aggregate and synthesize high-volume AI news and research datasets in real time.
• Architected a scalable RAG-based search and recommendation system for the platform's AI tools directory, enabling real-time semantic retrieval across 10,000+ AI products; served 500+ users, driving an 18% increase in CTR and generating $1M in incremental annual revenue.
• Engineered high-throughput inference workflows via dynamic batching strategies, slicing P95 latency by 7% while sustaining 3,000+ inference requests per second (RPS) across distributed serving environments.
• Optimized AWS EC2 cluster utilization to drive a 15% increase in infrastructure efficiency, leveraging TensorRT acceleration, embedding caching, and asynchronous request scheduling across high concurrency personalization engines.
LLM AI Agent Python RAG MLOps
Damco Solutions (AI &ML Services)
Machine Learning Engineer
October 2022 - April 2025 (2 years and 6 months)
Plainsboro Township, New Jersey, United States
• Co-designed and developed a multimodal healthcare diagnostic recommendation engine integrating LLMs with classical ML algorithms, achieving 93% data extraction accuracy backed by a Human-in-the-Loop (HITL) validation framework.
• Engineered high-performance Retrieval-Augmented Generation (RAG) pipelines for economic analysis, leveraging advanced embedding strategies, optimized retrieval, and semantic reranking achieving 90% data accuracy and reducing manual research time by 60%.
• Architected distributed PySpark data pipelines on AWS EMR and S3, orchestrating large-scale embedding generation and catalog feature engineering for millions of marketplace products.
• Deployed end-to-end MLOps workflows using AWS SageMaker and Apache Airflow, integrating Jenkins CI/CD pipelines to automate LLM deployment and lifecycle management while maintaining 99.9% production uptime.
• Led migration of legacy ML models to containerized architecture, improving deployment speed by 4x and reducing infrastructure costs by $120K annually.
LLM Docker AWS Fargate Jenkins RAG
IT Future Farm
Machine Learning Engineer
August 2018 - March 2022 (3 years and 7 months)
Chișinău, CV, Moldova
• Led the development of real-time computer vision systems for robotics and retail applications; engineered optimized edge-deployment pipelines that achieved over 96% detection accuracy.
• Engineered an automated document intelligence system using AWS Textract and NLP, cutting manual auditing time by 80% while scaling capacity to process 1000 financial documents daily at a 95% accuracy rate.
Computer Vision C++ NLP Python MLOps