K
Karli Sabalisk
About
Detail
California, United States
Résumé
Jobs
verified_user
0% verified
SeniorAI/ML EngineerN-iXApr 2024 - Sep 2025 (1 year 6 months)• Architected large-scale, cloud-native AI pipelines using Python, PyTorch, and TensorFlow for predictive analytics and anomaly detection across manufacturing systems. • Designed and deployed distributed inference microservices via FastAPI and Azure Functions, achieving ultra-low latency (<100ms) for real-time telemetry processing. • Developed high-throughput data ingestion and feature engineering pipelines using Azure Databricks, Delta Lake, and Kafka to handle millions of IoT signals daily. • Implemented automated retraining workflows and experiment tracking using MLflow and Azure ML, ensuring reproducibility and efficient model lifecycle management. • Applied SHAP and LIME frameworks for model explainability, integrating interpretab- Senior AI/ML Engineer | Software EngineerDedicattedMay 2022 - Mar 2024 (1 year 11 months)• Engineered multilingual OCR + NMT pipelines combining layout detection, document segmentation, and neural translationmodelsforglobaldocumentworkflows. • Built scalable microservices leveraging PaddleOCR, Tesseract, and Hugging Face Transformers to process large documentvolumesinreal-time. • Designed modular data extraction APIs integrated into a microservice-based SaaS architecture using Docker and Kubernetes. • ImprovedNLPmodelthroughputbyoptimizingbatchingandcachingmechanisms,reducingtranslationlatencyby40%. • Ledtechnical migrationoflegacyOCRsystemsintodistributed, containerized infrastructurewith CI/CDautomationvia GitHubActions. • Coordinatedwith productteamstotranslatetechnicaladvancementsintoimproved businessintelligencepipel
- CArtificial Intelligence EngineerCeladonsoftJan 2019 - Apr 2022 (3 years 4 months)• Developed AI-driven vulnerability detection tools that analyzed Solidity smart contracts through AST-based code parsingandGPT-poweredsemanticreasoning. • Designed autonomous multi-agent auditing systems leveraging LangChain, Model Control Protocol (MCP), and LLM coordinationforreal-timethreatdetection. • Built hybrid NLP + blockchain intelligence models to identify anomalies in on-chain behavioral data, improving detectionofsmartcontractexploits. • Implementedared-teamingsimulationlayertoevaluateand hardentherobustnessofdeployedLLMsystems. • Delivered modularAI components integrated with existing DevSecOpsworkflows, enabling fasterand safercontract analysispipelines. • Published internal documentation and architectural blueprintsout
- MSenior Machine Learning EngineerMindTitanFeb 2016 - Dec 2018 (2 years 11 months)• Designed AI-powered solutions forfraud detection, recommendation engines, and customer behavioranalysis using PyTorchandTensorFlow. • CreatedrobustAPIswith Flaskand FastAPItoserveMLmodelsinproductionwithsub-200msresponselatency. • BuiltNLPpipelinesleveragingspaCy, BERT,and Hugging FaceTransformersfortextclassification,sentimentanalysis, and namedentityrecognition. • Implementedend-to-end MLOpspipelinesonAWSandGCPwith Dockerizedworkflowsandcontinuousintegrationvia MLflow. • Collaborated with product owners and stakeholders to embed predictive intelligence directly into customer dashboards, driving measurable ROI gains.
- SPython Backend DeveloperSimform—Oct 2013 - Dec 2015 (2 years 3 months)• Built Python/Django-based backend systems with PostgreSQL, integrating early ML models for automated classification and analytics. • Designed asynchronous task queues using Celery and Redis, reducing response times by 25% for data-heavy operations. • Supported ETL workflows to extract and clean large datasets fordownstream analytics and BI reporting. • Collaborated with senior developers to implementAPI versioning, unit testing, and continuous deployment pipelines.
- SSoftware Engineer• Developed backend services in Python (Flask, Django) and integrated RESTful APIs for real-time communication betweenAI modulesandfront-enddashboards. • Implemented data-driven microservices supporting model inference, document processing, and structured output generation. • Contributed to reusable data transformation libraries in PySparkand Pandas, standardizing pre-processing pipelines acrossAI products. • Enhanced monitoringandobservabilitywithPrometheusandGrafana,enablingproactiveperformancemanagement. • Collaboratedwith MLOpsengineersto improvedeploymentautomation and reduce release cyclesthrough Terraform and Dockeroptimization. • Supported model validation and continuous improvement by integrating test-driven frameworks and au
Projects (professional or personal)
verified_user
0% verified
- SSmart Contract Vulnerability Analyzer
- RReal-Time Liveness Detection
- AAgentic Healthcare Assistant
- SSecure GenAI Workflow on GCPBuilt an end-to-end LLM pipeline using Vertex AI + AutoML with Model Armour to protect model integrity and monitor prompt safety. Designed autonomous LLM agents via MCP + LangGraph to assist patients with documentation and automated form filling. Implemented a FastAPI-based CV system using head-pose estimation and facial embeddings, reducing fraud risk in eKYC by 40%. Developed an LLM-driven code auditor using GPT-based reasoning and vector similarity search to detect blockchain security flaws.