Suhani Patel

Full-stack engineer with 7+ years delivering production SaaS platforms across education, automotive, and simulation domains using .NET Core, C#, Azure/AWS, and modern front-end frameworks including Blazor and React. Led the architecture and development of a GPT-Vision-powered passport/document-reading pipeline for a live university admissions workflow, reducing manual data entry by 70%. Experienced in modernising legacy applications and integrating AI/LLM capabilities into real-world, production systems.

Usually works in small full-stack teams, owning features from API design through backend logic to the final UI.

Software Engineer (Contract) at Magic Square Systems Limited (2024-09 – 2026-04)

Delivered client SaaS platforms for Coventry University, automotive dealership groups, and simulation/training providers, as part of a 2-3 person delivery team — independently delivering several features end-to-end while collaborating closely with others on the rest.

• Built two GPT-4 Vision document extraction pipelines for Coventry University's PGR+ research application portal. One reads an applicant's passport to identify their country and auto-fill nationality. The other reads education certificates and transcripts to pull out course title, start date, end date, and grade. Both auto-fill the application form and fall back to manual entry when extraction fails, which cut manual review for international applicants by 70%.
• Replaced a legacy FTPS/CSV stock feed with a token-authenticated Autotrader API integration across four dealership brands (Midlands Truck & Van, Bell Truck & Van, Used Vans 4 U, Used Trucks 4 U). Vehicle stock syncs in real time, with records updated, added, or removed as Autotrader's data changes.
• Migrated Predict Media from ASP.NET MVC to.NET Core by embedding Blazor components directly into the existing application. Kept the existing database, session handling, cookies, and authentication intact, so the transition had zero downtime and logged-in users saw no disruption.
• Built a centralized notification/activity-logging service, injected into other services via Dependency Injection for consistent use across the platform.
• Built an AI bot system for Predict Media's social platform. Bots are generated on demand, with a DALL·E-generated profile picture matched to a region the operator selects. Once created, bots post on topics the operator selected and can like, comment on, and repost other players' content in-character. Operators monitor and moderate all of this from a 'My Social Posts' dashboard.
• Added a scheduled job (trigger-based, runs every 10 minutes) that checks whether bot is enabled or disabled, so operators don't have to check bot status manually.
• Built a QR-code check-in flow for The Simulation Centre that automatically allocates delegates to cabins on scan, cutting check-in time by 40–60%.
• Rolled out Google reCAPTCHA v3 and bot-prevention checks as a standard approach across multiple client forms. Also added client-side image cropping and compression for Predict Media, cutting upload sizes by up to 50%.
• Implemented AWS S3 storage for secure media upload, retrieval, and lifecycle management across client platforms.
• Prototyped an automated WhatsApp notification workflow as a proof-of-concept, built using GitHub Copilot agent mode. A SQL Server trigger fires on new record insertion, looks up the assigned user's contact details, and sends a real-time alert via the Twilio API.
• Delivered features across 14 client platforms spanning education, automotive, and simulation sectors, including CU Ethics, OMIS Future Travel, HOPE, and CES Portal, with consistent ownership on the higher-depth projects above.

Software Developer at Array Bit (2019-09 – 2022-06)

• Built the contract expiry notification system for Benchmark Broker System, a client/customer contract management application: an Azure-scheduled job runs daily checks each morning and automatically emails customers at their chosen notice period (30, 60, or 90 days) before of contract expiry.
• Developed and maintained enterprise web applications for external clients — including Procure Safe, Atlica, and Dhivar Samaj — using.NET Framework, Angular, and SQL Server, working largely independently across the full stack.
• Optimised stored procedures and data access layers, reducing recurring data-access bugs and improving application reliability.
• Refactored and modularised legacy codebases, improving long-term maintainability and making it easier for new developers to onboard onto existing projects.
• Contributed to code quality by mentoring junior developers and writing unit tests using xUnit and NUnit, improving reliability and reducing regressions across the codebase.

Junior Software Developer at Natrix software Private Limited (2016-06 – 2019-07)

• Built DPRS, an internal tool for developers to log daily task activity — project, task, difficulty, and status — giving managers visibility into team workload and progress without manual check-ins.
• Built ASP.NET MVC applications within a 3-tier architecture (presentation, business, and database layers), contributing to core feature development and ongoing maintenance.
• Developed REST APIs and SQL components using ADO.NET and LINQ, and worked on stored procedures, queries, and functions as part of database design and development.
• Improved UI/UX using jQuery, AJAX, and.NET Validation Controls, building more interactive and responsive forms.
• Took part in code testing and bug resolution across the full software development lifecycle, from requirements through deployment.

MSc in Computing – Wrexham Glyndŵr University (2022 – 2024)

Master of Computer Applications (MCA) in Computer Applications – Uka Tarsadiya University (2014 – 2016)

Bachelor of Computer Applications (BCA) in Computer Applications – Veer Narmad South Gujarat University (2011 – 2014)