Looking for a job

Parag Nirwan

@paragnirwan

Message

AI and Full-Stack Engineer at Previsec who built an on-premise crisis decision-support system with local LLMs and RAG.

Germany

Message

What I'm looking for

I want to turn agentic AI into reliable enterprise solutions for asset-intensive industries.

At Previsec, I built an on-premise crisis decision-support system for the Previsec crisis-management product. I evaluated local language models and created a constrained RAG pipeline aligned to the FORDEC crisis framework.

At gotomaxx Software GmbH, I built an archive application end-to-end, including document search, retention management, and integrations with identity providers. I also made multi-tenant data-layer changes across Entity Framework Core, PostgreSQL, and SQLite.

Earlier, at Lensing.AI, I improved an ERP Applicants page from approximately 16 seconds to under 2 seconds. I recently built NewsLens, a Python AI research agent that uses bounded web research, source filtering, and claim tracking; I want to turn agentic AI into reliable enterprise solutions for asset-intensive industries.

Experience

Work history, roles, and key accomplishments

Full-Stack Software Engineer

gotomaxx Software GmbH

May 2026 - Sep 2026 (4 months)

Built an archive application end-to-end with .NET 10, ASP.NET Core, React 19, and TypeScript, covering data modeling, APIs, business logic, and UI. Implemented document previewing, faceted search, retention management, automated thumbnail generation, and staged ZIP exports for bulk document downloads.

.NET ASP.NET Core React TypeScript Entity Framework Core PostgreSQL SQLite IMAP Microsoft Entra ID Google Workspace OAuth OAuth 2 REST APIs Docker CI CD

UI and Software Engineer (Werkstudent)

Medical Equipment Europe

Nov 2024 - Sep 2025 (10 months)

Upgraded core application framework from .NET 5 to .NET 9 and migrated 15+ legacy WinForms views to WPF (C#/XAML) for medical software, creating a reusable UI component library (MVVM). Architected a modular setup module and navigation hierarchy, integrated native PDF preview and printing, and refactored legacy rendering code to lower memory footprint by 5%.

.NET C#XAML WPF MVVM Agile Performance Optimization

AI and Full-Stack Engineer (Master’s Thesis)

Previsec

Jan 2025 - Jul 2025 (6 months)

Architected an on-premise crisis decision-support system evaluating 4 local SLMs (Llama 3.2, Mistral, Gemma, Falcon) via Ollama, benchmarked across 5 NLP metrics. Created a multi-tenant AI service in TypeScript, Node.js, PostgreSQL, and Redis, building a constrained RAG pipeline with Crawl4AI scraping and Llama 3.2-Vision image context aligned to the FORDEC crisis framework.

TypeScript NodeJS PostgreSQL Redis RAG Ollama Llama 3.2 Mistral Gemma Falcon Docker GitLab CI CD NLP ROUGE BERTScore Perplexity

Application Engineer (Werkstudent)

Lensing.AI

Jul 2023 - Sep 2024 (1 year 2 months)

Developed backend microservices in TypeScript, JavaScript, and Node.js for a cloud-based ERP platform, implementing new business features and data-management functionality. Optimized the ERP Applicants page by identifying and resolving a data-retrieval bottleneck, reducing page load time from approximately 16 seconds to under 2 seconds (>87% improvement).

TypeScript JavaScript NodeJS PostgreSQL REST APIs Jest Unit Testing Performance Optimization

Operations Executive (IT Support)

Infosys Ltd.

Oct 2020 - May 2022 (1 year 7 months)

Provided L3 application support for enterprise applications built with Java, Spring Boot, and Angular 8, investigating and resolving production issues. Debugged and modified backend Java/Spring Boot and frontend Angular 8 code to identify defects and implement fixes.

Java Spring Boot Angular Jira Confluence Incident Management