AI That Plays Nice: Building a Privacy-First Grant Review Tool in Django

Speaker: Lionel Jouvet 16:45 stage 🎤

How do you actually bring LLM-powered analysis into a Django app without giving up control, privacy, or trust? We built a real production tool for university research administrators, and this talk walks through how we did it: dedicated worker processes for long analysis jobs, Server-Sent Events for live progress updates, LLM integration that works with local or or EU providers (Mistral), solid PDF section detection, and AI output that’s editable, not gospel. It’s a practical look at pairing Django’s strengths with modern AI tooling, the responsible way.

Abstract

At SDU, we built a Proposal Feedback Tool to help research-support staff and scientists get fast, consistent AI-assisted feedback on grant proposals. The catch? We didn’t want to lose human judgment, institutional context, or control over where the data lives. This talk is about the Django architecture that made that balance possible.

Here’s what we’ll dig into:

• Web/worker separation. Long-running LLM analysis doesn’t belong in a web request; it’ll just time out or lock things up. So we run it in a dedicated Django management-command worker instead. When someone reloads the page or comes back later, the browser reattaches to the in-progress job via Server-Sent Events. No duplicate work, no lost progress.

• Provider abstraction. Not every analysis needs the same horsepower. Sometimes local Ollama inference makes sense for privacy or cost. Sometimes you need the Mistral cloud API for better OCR or speed. We built it so each analysis can pick the right tool for the job.

• Resilient document processing. Grant proposals come in all shapes: some well-structured, some a mess. We combined section-detection heuristics with configurable pattern matching, added a verification UI so humans can double-check the machine’s work, and built a full-document fallback for shorter proposals that don’t need fancy parsing.

• Structured LLM outputs. AI-generated prompts, personas, and configs are only useful if they’re predictable. JSON mode and validation keep things from going sideways.

• Human-in-the-loop design. This is the part we care about most. AI output stays fully editable, but we always preserve the original result. Trust isn’t something you assume, it’s something you earn, one transparent interaction at a time.

• Schema-portable persistence. We built custom dump/load tooling on top of Django fixtures and natural keys, so backup and restore stays safe even as the schema evolves.

You’ll walk away with patterns you can actually reuse, whether you’re building for grant proposals, legal docs, or anything else where AI needs to assist, not replace, human judgment.

About Lionel

I grew up in a small village in the Jura, where my father was a wine producer, and have spent most of my career wandering between disciplines. I hold a PhD in Marine Biology, but over the years I have worked across ecology, conservation, bio-demography, microscopy, flow cytometry, eDNA, and a few other fields that seemed like a good idea at the time.

Along the way, I became a data scientist, working with everything from image analysis and natural language processing to spectroscopy, zoo databases (ZIMS), and IUCN Red List data. I have worked in both academia and industry and currently work at the University of Southern Denmark while living in Lyon, France. Through all of this, I have developed a healthy respect for the educational value of mistakes, having learned much of what I know by making them myself. Repeatedly.

I am also a co-founder of O Town Garage, a makerspace in Odense where I have taught 3D printing, CAD, design, sewing, and woodworking. I have been involved in 3D printing for more than a decade, long enough to remember when getting a print to finish successfully felt like a major scientific achievement. I am an active member of the open-source community, mostly by sharing designs and tools that others can build upon.

European at heart, I have lived in Scotland, Germany, Denmark, and now France, collecting experiences, friends, and hobbies along the way. My newest obsession is astrophotography. I have also built two Voron 3D printers named Corwin and Mini Yoda, which is slightly surprising given that I still cannot reliably tell left from right.

My professional philosophy is simple: expertise is often just a long series of mistakes, curiosity, and enough persistence to try again the next day. At this point, my career can best be described as an exceptionally long integration test that somehow keeps passing.

You can find most of my projects (academia productions, 3D things, astrophotographers, etc.) on my GitHub profile Ecolio.