N
Nicolas Rocha Guzman
About
Detail
Barranquilla, Atlantico, Colombia
Résumé
Jobs
verified_user
0% verified
Senior Python Software EngineerVeriskSep 2024 - Current (1 year 9 months)• Architected a comprehensive predictive analytics platform for dynamic insurance premium modeling, orchestrating end-to-end data ingestion pipelines from disparate telematics and claims sources to deliver real- time risk assessments for millions of policyholders while ensuring sub-millisecond query latencies under peak loads. • Engineered core backend services using Python and FastAPI to power the inference layer, structuring asynchronous handlers that processed high-velocity streaming data and reduced model deployment cycles from weeks to hours. • Implemented machine learning pipelines with scikit-learn and TensorFlow, training ensemble models on terabyte-scale datasets to achieve 25% improvements in predictive accuracy for catastrophe ri- RSenior Python Software EngineerRoundTechSquareMay 2022 - Aug 2024 (2 years 4 months)• Directed the development of an intelligent fraud detection engine for financial transaction processing, redesigning the feature engineering layer to incorporate temporal graph analytics and handle billions of daily events with fault-tolerant distributed computing. • Developed robust data processing pipelines in Python integrated with Pandas and Dask, enabling parallel computations that scaled horizontally to ingest and transform petabyte volumes of heterogeneous transaction logs. • Built computer vision models using PyTorch, fine-tuning convolutional networks on labeled imagery datasets to detect synthetic fraud artifacts, achieving 30% uplift in recall rates over legacy rule-based systems. • Orchestrated containerized deployments on Kube
- Python Software EngineerPolluxaSep 2020 - Apr 2022 (1 year 8 months)• Spearheaded the creation of a geospatial analytics dashboard for intelligence data fusion, architecting modular pipelines that correlated satellite imagery with ground sensor feeds to generate actionable threat visualizations for operational teams. • Constructed backend APIs in Python with Flask, handling geospatial queries and vector operations that powered interactive mapping layers serving thousands of concurrent analysts. • Developed classical ML algorithms using scikit-learn, clustering multivariate sensor data to identify anomalous patterns, which enhanced early-warning capabilities by 35% in simulated scenarios. • Migrated monolithic services to Docker containers orchestrated on AWS ECS, incorporating CloudWatch Logs and X-Ray for
- NPython Software EngineerNexGenOct 2018 - Aug 2020 (1 year 11 months)• Led the engineering of a supply chain optimization platform for logistics forecasting, designing simulation engines that modeled multi-echelon inventory flows under stochastic demand variability to minimize stockouts across global distribution networks. • Implemented simulation models in Python utilizing NumPy and SciPy, performing Monte Carlo analyses that quantified risk exposures and informed strategic replenishment policies. • Deployed optimization solvers with Gurobi, solving mixed-integer linear programs for route planning that delivered 20% reductions in transportation costs for enterprise-scale fleets. • Established CI/CD pipelines using Jenkins on AWS CodePipeline, automating builds and tests that ensured regression-free releases
- Junior Python Software DeveloperZigronJul 2016 - Sep 2018 (2 years 3 months)• Built foundational data processing tools for geospatial intelligence workflows, creating batch ingestion scripts that standardized formats from various satellite and aerial sources for downstream analytic consumption. • Programmed core utilities in Python with NumPy, vectorizing image preprocessing routines that accelerated alignment and mosaicking tasks by 50x on commodity hardware. • Integrated open-source libraries like GDAL for raster manipulation, enabling format conversions and reprojections that unified disparate datasets for fusion algorithms. • Deployed applications on EC2 instances with Auto Scaling Groups, configuring ELB load balancers to distribute workloads across availability zones reliably. • Optimized database schemas in
Education
verified_user
0% verified
- Bachelor of Science | Computer ScienceCarleton UniversityJan 2012 - Jan 2016 (4 years 1 month)