Procore Technologies — Full Stack Developer, Estimating

TypeScript, React, MobX, C#, ASP.NET Core, EF Core, SQL Server, REST, Kubernetes, Docker, AWS S3, Playwright, Jest

• Designed and built ~20 endpoints of the Esticom Public API v2.0 (estimate summary resources) from scratch in 5–6 weeks: API contract, PUT/PATCH semantics, validation, controllers, and data access. Now used by external customer integrations.
• Built an LLM-based import pipeline for CSV/XLS estimates with semantic column mapping and a human-in-the-loop fallback for low-confidence matches, eliminating manual column mapping on import.
• Delivered a customer-requested redesign of the estimate summary table: flattened the nested group › assembly › part data from the MobX store and extended the React table with per-level visibility toggles and configurable VAT/overhead totals.
• Added new column types to the estimating grid (a large virtualized data grid with inline-editable cells), including S3-backed file-attachment columns with presigned-URL upload/download, with the corresponding API and React UI.
• Set up the Kubernetes-based test environments used by the team's Playwright and integration suites; containerized services for both standalone and integrated configurations.
• Piloted AI-assisted development on a full epic; wrote the report the team adopted as its workflow reference.

Novanta — Software Design Engineer

C#, .NET, WinForms, NUnit, embedded C++, Linux, Azure DevOps, IEC 62304

• Authored the product's state-machine specification (device signals, application states, UI interactions) for a regulated medical pump device; used as the software contract in design reviews and test planning under an IEC 62304 process.
• Implemented the layer translating hardware signals into deterministic LED and UI states (ready / normal / high-pressure), including the application–driver interface.
• Built a reusable C# tab-navigation component library adopted across all tabbed device screens, replacing per-screen duplicate implementations.
• Served as Agile Team Coach for the 11-person team: ran retrospectives and coordinated across hardware, software, and QA. Maintained the NUnit test pipeline in Azure DevOps.

Research positions — 2011 – 2023

Postdoctoral researcher, doctoral researcher, and research assistant in algorithms and computational geometry, with teaching and nine peer-reviewed publications. Details on the Academia tab.

Postdoctoral Researcher — Toronto Metropolitan University

• Designed and implemented motion-planning algorithms for minimal-sensing robots in Steven LaValle's group; two peer-reviewed publications.
• Built a polygon decomposition and "smart bouncing" strategy for sensor-limited robots, validated against reflective and random strategies across 500 simulations.
• Lectured Algorithms (CPS616), a 120-student undergraduate course, for a semester; mentored students; helped organize CCCG 2022.

Doctoral Researcher — Masaryk University

• PhD on geometric graph recognition and complexity; seven peer-reviewed publications in graph algorithms and computational geometry.
• Formal complexity analysis and proofs for graph-theoretic algorithms in combinatorial optimization problems.
• Mentor for the Graph Theory course over several semesters.
• Research visits to ENS de Lyon, Jagiellonian University, UNAM, Cinvestav, and UASLP.

Research Assistant — Izmir University of Economics

• Designed and implemented a multi-stage 3D localization algorithm for wireless sensor networks in Java, combining plane-level self-localization with inter-plane alignment using linear algebra.
• Developed and benchmarked vehicle routing heuristics for a TÜBİTAK-funded logistics project, comparing custom algorithms against optimal solutions and randomized baselines in Python and C++.
• Assisted courses in Data Structures, Algorithms, Database Systems, and Discrete Mathematics over four years.