Ravi Sharma, Senior MLOps Engineer

Hi, I’m Ravi Sharma

Approximately 11 Years of Experience

Scalable Solutions, Infinite Possibilities

Architecting secure, scalable solutions for the future.

Discover My Work

About Me

Over a decade ago, I started as a Junior IT Executive at Cafe Coffee Day, troubleshooting networks and scripting solutions. That curiosity evolved into a passion for automation and innovation.

From managing servers at Acma Computers to leading MLOps at Nityo Tech, Avatar Techno Services, and SPH Media in Singapore, I’ve turned complex challenges into streamlined systems. Today, I deploy AI models, secure cloud infrastructures, and integrate IoT, embracing generative AI and adversarial robustness.

Where I Want to Be: In 5-10 years, I aim to lead global MLOps initiatives, designing secure, scalable AI and IoT frameworks. I envision mentoring the next generation of engineers and driving enterprises toward a future where technology empowers without compromising safety. Looking ahead, I’m excited to dive into quantum computing—exploring its implementations for AI acceleration, understanding its paradigm shift, and tackling security challenges like quantum-safe cryptography. I also aspire to pioneer hybrid classical-quantum MLOps pipelines and innovate in edge AI security, pushing the boundaries of what technology can achieve.

What Sets Me Apart: A problem-solver with a positive, action-oriented mindset, I bring velocity and reliability to every project, and fueled by a love for chess, badminton, and cutting-edge tech. Colleagues call me ‘cool and punctual’.

Projects

Generative AI Security Pipeline

Developed a secure MLOps pipeline for generative AI models like Stable Diffusion on AWS SageMaker and GCP Vertex AI with robust defenses.

Challenges: Prompt injection bypassed filters, and model inversion risked data exposure.

Solution: Used ART and Foolbox for adversarial training, encrypted inputs with AES-256 via AWS KMS, monitored with Splunk.

Outcome: Cut attack success by 85%, boosted resilience by 30%, ensured HIPAA/NIST compliance.

Ethical Hacking of Financial Web App

Conducted ethical hacking and penetration testing on a financial web app to identify and mitigate critical security flaws.

Challenges: Encrypted traffic hid vulnerabilities; weak authentication enabled privilege escalation.

Solution: Decrypted SSL with Proxychains, exploited flaws using SQLMap and Mimikatz, recommended MFA enhancements.

Outcome: Found 15 vulnerabilities including 3 escalations, hardened security, achieved PCIDSS compliance.

Secure IoT for Smart Manufacturing

Built a secure IoT system for a manufacturing plant using Node-RED and AWS IoT Core for real-time PLC sensor data.

Challenges: Sensor latency delayed updates; open ports risked unauthorized access.

Solution: Optimized data flow with Node-RED, secured with TLS and Sysdig, scanned vulnerabilities with Nmap.

Outcome: Achieved <1-second latency, blocked 98% intrusions, met GDPR standards.

Big Data Fraud Detection Automation

Automated a fraud detection pipeline processing 10TB of data with PySpark, Snowflake, and AWS Glue, integrating ML models.

Challenges: Data latency slowed alerts; resource overuse increased costs.

Solution: Optimized ETL with Glue, scaled Spark clusters, monitored performance using Prometheus.

Outcome: Reduced processing time by 60%, cut costs by 25%, detected 95% fraud.

Scalable ML Deployment Pipeline

Created a CI/CD pipeline for ML models with Jenkins, Kubernetes (EKS), and MLflow for automated deployment efficiency.

Challenges: Model drift degraded performance; scaling issues caused downtime.

Solution: Tuned Kubernetes resources, monitored drift with MLflow, automated retraining workflows.

Outcome: Slashed deployment time to 4 hours, achieved 99.9% uptime.

Multi-Cloud Backup Automation

Designed an automated backup system across AWS, GCP, and Azure using Terraform, Ansible, and Python for reliability.

Challenges: Inconsistent schedules risked data loss; key rotation disrupted access.

Solution: Scripted cron jobs with Terraform and Ansible, secured keys with HashiCorp Vault, monitored via Grafana.

Outcome: Ensured 100% backup reliability, sped recovery by 70%, GDPR compliant.

Skills

MLOps & AI Automation

TensorFlow, PyTorch, Gemini, LLM, MLflow, Vertex AI, SageMaker, Stability AI, Gradio, Hugging Face, Terraform, Ansible, Helm, ArgoCD, Grafana, Splunk, SSO (Okta, Auth0)

DevOps & CI/CD Pipelines

Docker, Kubernetes (EKS, GKE, AKS),OpenShift, Load Balancer, Jenkins, GitLab, GitHub Actions, Bitbucket, Azure Pipeline, CodePipeline, Octopus Deploy, XL XebiaLabs, HashiCorp Vault

Cloud Platforms

AWS (ECS, Lambda, S3, IAM, ECR, VPC, SG, SM, RDS, CloudFormation, Autoscalling, VPC, Route53, EventBridge), GCP, Azure (Key Vault, Databricks, Functions, ARM), CDN, IBM Cloud, Alibaba Cloud

MLSecOps & Cybersecurity

ART, CleverHans, Foolbox, Qualys, Sysdig, Checkov, Cryptography, Burp Suite, OWASP ZAP, SAST, DAST, WAF, SIEM,Tenable; HIPAA, NIST, GDPR, PCIDSS compliance

Big Data & IoT

Hadoop, PySpark, Snowflake, EMR, Glue, Athena, BigQuery, Redshift, Node-RED, PLC Sensors integration, Kafka, Pub/Sub, RabbitMQ, Data Lake

Programming & Dev Tools

Python, Java, Node.js, Golang, Vue.js, Ruby, Bash, PowerShell, Maven, Gradle, Postman, Selenium, Jira, Traefik, Apache, Nginx, Istio

Databases & Backup

PostgreSQL, MySQL, MongoDB, DynamoDB, Oracle, TSDB, Backup Solutions (AWS, GCP, Azure), Kasten K10, Veeam, Teradata

Ethical Hacking & Pen Testing

Nmap, Metasploit, Wireshark, Mimikatz, Red-Hawk, SQLMap, Nessus, Burp Suite, OWASP ZAP, Proxychains, Hashcat

Training & Certifications

Contact Me

Email me at rlucys@hotmail.com or connect on LinkedIn.