CV

04.1 / experience

Experience

Travel Cue Management (travel tech), Aug 2021 – now: four roles in five years, in Mumbai and, since Aug 2024, Hamburg (remote).

• Aug 2024 – now

  Software Developer (part-time, 20 h a week, alongside the M.Sc.)

  Travel Cue Management · Hamburg (remote)

  • Cut fare processing in the SBT application from over 5 minutes to 4–5 seconds for 250+ fares per request, each with fee and settlement calculation. Done through caching, SQL query and stored-procedure optimisation, and in-memory dictionary lookups.
  • Designed a central FareEngine that unified fare calculation previously duplicated across the POS booking application and the Cue accounting application, fixing long-standing calculation defects and backing the logic with a unit-test suite.
  • Redesigned the Compas B2B integration to reuse the applications that own each domain (CRM for traveller records, POS for request handling), removing mapping logic duplicated across several internal applications.
  • Migrated legacy applications in the core suite to API-driven architectures, and built full-stack RezCue features (React + TypeScript interfaces backed by .NET APIs) used daily by the operations team.
• Oct 2023 – Mar 2024

  Software Developer

  Travel Cue Management · Mumbai, India

  • Built an automated passport-scanning API that made traveller creation and search available directly from the request workflow, replacing manual data entry and saving 4–5 minutes per travel request.
  • Compared OCR providers on accuracy and cost, and led the choice for production use.
  • Implemented identity idempotency logic that prevented duplicate traveller profiles and the manual clean-up they caused.
  • Developed responsive interfaces in React, TypeScript and Bootstrap 5 for the traveller-profile workflows.
• Feb 2022 – Sep 2023

  Junior Software Developer

  Travel Cue Management · Mumbai, India

  • Delivered full-stack work across API projects and .NET Core services, including SQL query optimisation and complex features, and mentored new joiners.
  • Led the Cytric integration end to end and gave backup support to the front-end team.
  • Built proofs of concept for auto-generated ICS calendar files, Power BI reporting, HR-system integration via merge.dev, and external partners including Thomalex.
• Aug 2021 – Jan 2022

  Trainee Software Developer

  Travel Cue Management · Mumbai, India

  • Shipped features and defect fixes in JavaScript, C# and .NET Framework, and built a Blazor proof of concept for the AdminTool application.

04.2 / projects

Projects and open source

• Oct 2026

  Mullion: open-source macOS window manager

  Swift, AppKit, SwiftUI

  • Built and released v1.0 of a native menu-bar app with keyboard-shortcut and drag-to-edge window snapping, multi-monitor moves and configurable settings.
  Mullion case study
• Oct 2026

  42nd Cloudflight Coding Contest

  Team of two

  • 2nd at Universität Paderborn, 5th nationally and 84th worldwide.
  LinkedIn post
• Sep 2026

  LeezenPass: MÜNSTERHACK 2026

  ASP.NET Core, React, PostgreSQL, PostGIS

  • Proposed the idea, worked on design and code, and gave the pitch for a bike-theft prevention app built with a team of five in 36 hours: digital bike pass, public frame-number check, theft reports, ownership transfer with one-time codes, and AI prefill from photos, with the model running locally. 59 audience votes.
  LeezenPass case study
• 2026

  Open-source contributions

  C#, TypeScript, .NET and developer tooling

  • Pull requests to anthropics/anthropic-sdk-csharp, dotnet/extensions, orval-labs/orval, dotnet/roslynator, TNG/ArchUnitNET, and modelcontextprotocol/csharp-sdk.
  All pull requests
• 2026

  Ankommen in Paderborn

  Astro, React, TypeScript, Tailwind CSS

  • A free, unofficial guide in English and German for international students arriving in Paderborn, in which every fact links to an official source. Twelve topics so far; first release planned for December 2026.
  Ankommen in Paderborn case study

04.3 / education

Education

• Apr 2024 – now

  M.Sc. Computer Science

  Universität Paderborn

  • Master’s thesis: “Advancing Architectural Tactic Detection through Large Language Models”, supervised by Jun.-Prof. Dr. Mohamed Soliman, second supervisor Prof. Dr. Eric Bodden. Thesis due Feb 2027.
  • Building an LLM pipeline that detects architectural tactics in Java code changes from open-source projects (e.g. Apache Hadoop); evaluating IBM Granite 4.2 and Qwen3-Coder (30B) with zero- and few-shot prompting on the university’s HPC cluster.
  • Face-recognition demonstrator: React + TypeScript interface for a QMagFace pipeline, shown at the Heinz Nixdorf MuseumsForum.
• Jun 2018 – Mar 2021

  B.Sc. Computer Science

  Ramnarain Ruia Autonomous College, Mumbai

04.4 / skills

Skills

programming

C#, TypeScript, JavaScript, SQL, Python, Java

backend

.NET Core, .NET Framework, ASP.NET, REST API design, Entity Framework, Blazor

frontend

React, TypeScript, Tailwind CSS, Bootstrap 5, HTML/CSS

cloud and devops

Docker, Azure, Git, CI/CD

ai / llm

LLM pipelines in Python, prompt engineering (zero- and few-shot), LLM evaluation, open-weight models (IBM Granite, Qwen3-Coder), HPC/Slurm, AI-assisted development (Claude Code)

data and performance

SQL Server, query and stored-procedure optimisation, caching, profiling, data modelling

architecture and practices

DDD, SOLID, architectural patterns, quality attributes, architecture recovery, unit testing, code review

security and integrations

OpenID Connect (OIDC), SAML, web vulnerability mitigation; Cytric, Compas, Thomalex, merge.dev, OCR APIs

languages

Hindi (native), English (C1), German (A2)