Adrian Paleacu

Senior MLOps Engineer & AI Platform Architect

adrian.paleacu@gmail.com +40 721 910 503 Timisoara, Romania

I architect and build production ML systems that scale. With 20+ years in software engineering and deep expertise in MLOps since Kubeflow 0.4 (2018), I've designed AI platforms for Fortune 500 companies across healthcare, finance, identity, and telecom. I don't just deploy models - I build the entire ecosystem: from federated learning infrastructures to autonomous agentic coding environments. My systems handle 100x Spark workloads, serve multi-tenant enterprise deployments with advanced security clearances, and turn ML research into production-grade solutions.

20+
Years Experience
7+
Years in MLOps
20+
AI Interfaces Built

Core Expertise

Kubeflow Pioneer (since 2018) AI Platform Architecture Federated Learning KFServing / KServe Kubernetes & Operators MLflow & Databricks Nvidia Triton Python / FastAPI / gRPC Azure & AWS Istio Service Mesh Model Optimization (INT8/BERT) PCI DSS / OWASP Security

Innovation Spotlight

Current Project

Autonomous Agentic Coding Environment (AACE)

Designing the Future of AI-Assisted Development

Building a comprehensive system architecture for autonomous AI-powered software development. This platform orchestrates specialized AI agents for different development tasks, featuring a 3-tier memory architecture (long-term knowledge, pattern library, episodic memory) that enables continuous learning and improvement.

Multi-agent orchestration system
Knowledge base with vector search
Automated code generation & review
Policy guardian for AI safety
Ephemeral workspace management
Feedback loop for self-improvement
Vue 3 TypeScript Python FastAPI Kubernetes Vector DB MCP

Professional Experience

Senior MLOps Engineer / AI Solution Architect
Self-employed (Consulting)
Mar 2021 - Present
Modular AI Platform
US Professional Services Company
  • Architected and built 20+ integrated interfaces including Data Processing Dashboard, Federated Learning Dashboard, Model Deployment Console, and Agentic AI Workflow Designer
  • Designed custom Kubernetes Operators to manage lifecycle automation of ML components across environments
  • Delivered end-to-end solution covering UI/UX, backend, and infrastructure orchestration
VueJS TypeScript FastAPI gRPC K8s Operator SDK Flower.AI Argo CD
Edge Computing Platform for Data Science
Major Identity Provider
  • Developed cloud-agnostic platform handling 100x Spark workloads in Scala and Python
  • Built federated learning across distributed nodes - enabling collaborative model training without moving data
  • Implemented robust feature store and artifact storage ensuring data integrity and compliance
High-Scale Model Serving Infrastructure
Fortune 500 Computer Hardware Company
  • Designed model serving infrastructure on Azure using AKS, Databricks, MLflow, KFServing, and Istio
  • Architected solutions with multiple backends: MLflow, BentoML, MLeap, Nvidia Triton
  • Optimized BERT models using INT8 quantization for improved inference performance
  • Established deployment methodologies for multi-tenant environments with cyber-security clearances
AI Match Platform
Healthcare Workforce Development
  • Built AI-powered matching system for healthcare workforce training and credentialing platform
Senior MLOps Engineer / AI Solution Architect
Ness Digital Engineering
Jun 2016 - Mar 2021
Ness AI Accelerator Platform
  • Designed cloud-agnostic AI accelerator platform for rapid client onboarding
  • Developed AI reference architecture based on Google ML Rubric with model store, feature store, and MLOps best practices
  • Managed deployments on-premises and across Azure and AWS
  • Mentored engineers and conducted training sessions on MLOps tools
Kubernetes Kubeflow KFServing Istio CephFS Nvidia DeepOps
Real-Time ETA Prediction Platform
Government of Ontario, Canada
  • Deployed AI platform on Kubernetes for real-time ETA predictions
  • Built scalable infrastructure supporting model training, deployment, and production inference
Medical Imaging AI Platform
Varian (Siemens Healthcare)
  • Designed platform for organ segmentation from MRI and DICOM imaging
  • Built scalable architecture aligned with Google ML Rubric including model and feature stores
Kubeflow KFServing GlusterFS Azure ARM
AI Platform on OpenShift
Bell Canada (Telecom)
  • Defined full lifecycle AI platform integrating Spark with enterprise-grade security
Finance Big Data Solution
Warner Music Group
  • Built data pipelines processing millions of sales transactions and royalty payments
  • Designed scalable ingestion, transformation, and reconciliation for global financial reporting
Next-Gen IoT Architecture
GeoTogether (UK)
  • Led technical team designing event-driven microservices for IoT edge devices
  • Created real-time data streaming and cloud-native deployment architecture
Senior Software Engineer
IBM Global Technology Services Romania
Oct 2015 - Jun 2016
  • Developed software solutions for BMW automotive systems using Java EE, Eclipse RCP
  • Collaborated with international teams on the BNE product, improving performance and reliability
Senior Software Engineer
Paymint AG (Payment Industry)
Nov 2013 - Oct 2015
  • Engineered fraud prevention and dispute management systems for major German payment processor
  • Acted as Subject Matter Expert on PCI DSS and OWASP security compliance
  • Optimized database queries and application performance in Agile environment
Senior Software Engineer / Technical Leader
ACI Worldwide (Banking & Payments)
Nov 2006 - Nov 2013
  • Engineering Lead for ACI Case Management System built on semantic web framework (Apache Jena)
  • Led Proactive Risk Manager L3 Support Team serving 150+ banking clients
  • Modeled OWL 2.0 ontologies for knowledge representation
  • Undertook onsite assignments in USA and Switzerland
  • SME on PCI DSS and OWASP for enterprise stack (WebSphere, WebLogic)
Java Semantic Web Apache Jena OSGi WebSphere
Software Engineer
CoreCard (Banking & Payments)
Aug 2003 - Oct 2006
  • Built payment products: Card Management System, Collections, Sales Incentive Management
  • Contributed to platform handling Private Label, Prepaid, Fleet, and credit card payments

Certifications

Machine Learning Stanford University / Coursera (2016)
Data Science Specialization Johns Hopkins / Coursera (8 courses)
Hadoop Platform & Applications Coursera (2016)
Go Programming Specialization Coursera (3 courses)

Education

B.Sc. Computer Science

Tibiscus University, Timisoara (2000-2004)

Recognition

ACI Award of Distinction 2010
ACI Q2 AD Award 2008

Conferences & Events

AI World Summit 2018 (Amsterdam)  |  Google Cloud Days 2019 (Bucharest)  |  RAAI 2019  |  SYNASC 2019  |  Devoxx 2010 (Belgium)