Computer scientist Muskan Hashim doesn't just build AI models — her work helps support the infrastructure, standards, and governance they run on. She gained her computing expertise as a scholarship holder of the DAAD's Konrad Zuse Schools of Excellence in AI. She now works at Helmholtz Munich, in project management for Scverse, an open-source consortium for data analysis in the life sciences. Her goal: to help support reliable tools without which even the best AI research cannot function.

AI models rarely fail for lack of ideas — more often it's the environment they run in that lets them down: poorly maintained software, inconsistent data, missing standards. This is exactly what matters to Muskan Hashim. “Everyone wants to build groundbreaking AI models. Almost nobody wants to look after what they run on.” She sees this imbalance as one of the biggest reliability risks facing AI in the life sciences. For Muskan Hashim, reproducibility, data standards, and well-maintained tools are not a side issue — they are the crucial substance of reliable research.

At Scverse, that is precisely her job. Scverse develops freely accessible software tools for analysing biological data and operates across research labs all around the globe. The project was initiated at Helmholtz Munich, where Hashim now helps manage its organisational side. Researchers can use it to analyse data from individual cells — working out, for instance, which genes are active in a given body cell and how cells differ or change as a result (in diseases such as cancer, for example). 

For tens of thousands of scientists worldwide, this is an enormous help. Hashim works in Professor Fabian Theis's group at Helmholtz Munich — not just on any particular AI model, but on the infrastructure that allows the analysis of data that many models depend on in the first place. These tools are what researchers use to turn raw single-cell data into something an AI model can actually use and draw insights from.

Focused on Interdisciplinary and Applicable Research

After a bachelor's degree in Statistics at the University of Delhi, she worked alongside her studies and pursued a further diploma in programming. Wanting to work on projects with real impact drove her towards a master's in Scientific Computing at Heidelberg University. Unlike classical computer science, Scientific Computing is more application-driven and interdisciplinary: numerical methods, simulation, modelling, data analysis and, increasingly, machine learning are applied to concrete questions in the natural, life and engineering sciences.

It was there that she also learned about the Zuse School ELIZA. What drove her to apply was less the funding than a sense of alignment with the programme's values: “I looked at the vision and values of the Zuse Schools and felt a real affinity with them. That was my main motivation for applying.” The funding also gave her something more tangible: independence. “Being raised by a single mother who was both a strong role model and a source of inspiration, it always mattered to me to stand on my own two feet rather than depend on my family — I never wanted to rely on anyone else to fund my education.” Working alongside her studies is a matter of principle for her: “I've made it a rule never to just complete a programme, but to apply what I'm learning at the same time.” Through the Zuse School she collaborated with three research groups, and she still works on joint papers with them today. “In Heidelberg that runs through a mini job; with two further groups, I do research at weekends, in my spare time.”

Widening Access and Representation

Her research was interdisciplinary from the outset: epidemiology, the natural and social sciences — including work on the PANDASIA project at the Climate-Sensitive Infectious Diseases Lab, based at Heidelberg University's Interdisciplinary Centre for Scientific Computing (IWR). For her master's thesis, she analysed data to study the socio-behavioural drivers of zoonotic disease spillover risk. The underlying idea: everyone's individual behaviour contributes to the spread of infectious diseases, so changes in everyday behaviour could, in turn, help reduce that risk.

At IWR, Hashim also led Upstream, the network for women in STEM. Her motivation for this goes back to her childhood: in research, she says, there are restrictions everywhere that limit access to resources and opportunities — first it was barriers to education, later it was barriers linked to her gender. From this she draws her long-term goal: greater representation and better access to education. For her, open source is a natural consequence of that: the code is, in principle, accessible to everyone. She also brings strong industry experience from her time at a global Software as a Service (SaaS) platform — background that helps her navigate her current role at Scverse, where she works at the intersection of academia and industry.

Germany Offers Ideal Conditions

In the long run, Hashim wants to grow into technical leadership roles — while staying at the level where decisions have the greatest impact: standards, structures and platforms, rather than individual models. What she values most about the Zuse Schools' vision is precisely this commitment to reliable AI infrastructure: “AI should always be explainable — it's a tool we need to use and develop with a great deal of care.”

For her, Germany offers ideal conditions for this — for reasons that go beyond scientific excellence alone. It reflects values that matter to her: independence, efficiency, and a belief in science and technology as forces for good. “I didn't come here mainly to become part of a sought-after academic community, but because this is where the work that matters to me is actually being done.”

Klaus Lüber (July 2026)