Overview
Helsens holds a PhD in biological and computational sciences from Ghent University. Before Solvice he worked as a Research Data Scientist at Biocartis and Sony, and later as ML Lead at In The Pocket. He co-founded Solvice with Christophe Van Huele in 2015, bringing a machine learning lens to the company's optimisation product stack.
Beyond Solvice, Helsens contributes to Climate Change AI.
Career history
In the news
- Next week Thu 29 Oct: Bits in Bio Show & Tell, how AI transforms your daily workflows. Slightly different format this time. A series of flash talks where people in our community show how they are putting AI agents to work in life sciences. Topics & titles I've seen on the schedule are covering, - Building novel UIs for a complex biological data workflows - World models in drug discovery - LLMs helping to build a biotech company brain - MCP integration with ChimeraX and PyMOL for molecular visualisation - Workflows and skills for
- Every series needs a first edition. Yesterday, our CTO Quinten Rosseel helped kick off the very first Thundertalks at Wintercircus Ghent: a new monthly session where builders get 5 to 10 minutes to show what they actually built and how. No product demos allowed, just builders teaching builders. Quinten Rosseel walked through Ella Energy AI’s agentic engineering stack and how Ella deals with the bottlenecks in its AI software factory. Now that AI agents write most of our code, building is no longer the slow part. The hard parts are
- Just submitted our work on challenge #2 of the Anthropic x Adaptyv/Proteinbase competition, target of the week TNF-alpha! 🧬 The brief this time: bind human TNF-alpha at pH 7.4, let go at pH 6.0, and recognise mouse TNF-alpha. A few numbers & learnings ↘️ 1/ Used our Crane Protein Intelligence Platform to generate 1402 miniproteins & nanobodies, yielding 44 strong candidates, further reducing to the final 20 designs for submission. The last bit from 44 -> 20 was pretty subjective, its the part where the scores saturate and human
- Worked on the Anthropic x Adaptyv/Proteinbase EGFR protein design challenge last week! 🧬 The brief: design a protein that binds EGFR-human more strongly at pH 6.5 than at pH 7.4, and also binds evolutionary related EGFR-mouse at pH 6.5 👨⚕️ 🎯 🐁 🧪 Our approach ↘️ 1/ ran the brief end-to-end on our Crane Protein Intelligence Platform, orchestrating Boltzgen/RFDiffusion/ESMFold2/PropKa/Chai-1/Protenix, iteratively analyzing and estimating metrics towards these objectives, qualifying epitope regions 2/ generated 2 arms of >1000
- I see a lot of people energized and empowered by AI around me, exploring how they can do in an hour what used to take a week, and all that AI unblocks for them. I also see that we're all figuring this out in our own way, learning what works, and where it still falls short big time. Being a community builder and learner at ❤️, I'm putting some extra time to help organizing two show & tell sessions in the in the next three weeks. No slides, just hands-on demo's and real people. 1️⃣ ➡️ Thundertalks at Wintercircus Ghent ⚡ Wednesday 14
- About two years ago, when working at VIB, I started lobbying/pushing to get more biotech & life sciences on the agenda of Wintercircus Ghent. So full circle on October 27, as VIB and Wintercircus Ghent are filling up the building with +400 tech and biotech people for FTI Tech x Bio Night - so this one is really close to my heart. Amongst others: Karen Massey ( argenx ), Roeland Delrue ( Aikido Security ), Riccardo Sabatini ( Orionis Biosciences ) and Jeroen Van Hautte 🐺 ( TechWolf ) will be sharing the stage and discuss
- More thinking on biosecurity. 💭 Published last year Science, ref in the bottom, Wittmann and colleagues tested more than 75,000 AI-redesigned protein variants against the screening tools DNA synthesis providers use to catch dangerous orders. The tools flag the original sequences almost perfectly. On the variants that were likely to keep their "dangerous" properties, many slipped through and could just be ordered. Before publication, three of four screening vendors patched their tools. Detection improved, but still 3% of
- Thinking on Biosafety. Ken Alibek was First Deputy Director of Biopreparat, the Soviet Union's biological weapons program, until he defected to the US in 1992. Most of the scientists working inside that program did not know that is what they were doing. Biopreparat ran on tiers of clearance. At the outer tier you were told your work defended against an American bioweapons program that did not exist. Only once you were promoted into an inner circle did anyone tell you the real goal was offense. Alibek wrote about this himself.
Related profiles
This page shows public professional information only, each fact cited. Is this you? send a correction, or ask for removal within 24 hours, no questions asked.