MOTIVATION LETTER The Journeys Incubator Programme opens in August 2026, and my venture is ready for that four-month sprint. I am Eniola Olutogun, a pharmacist turned machine learning engineer, and I have built a computational platform that predicts drug resistance mutations from protein sequence alone. The proof of concept is complete: AUROC 0.804 plus or minus 0.025 on the Platinum benchmark with 553 mutations under protein-grouped cross-validation, and 100 percent mutation coverage compared to roughly 18 percent for structure-dependent tools. The published state of the art, mCSM-lig, scores about 0.70 on the same benchmark. My classifier uses ESM-2 protein language model delta-embeddings combined with ECFP4 drug fingerprints and a Random Forest classifier. No crystal structure is required. Helsinki is a growing biotech hub, and the Journeys Incubator Programme offers exactly what this venture needs at this stage: a structured four-month window to convert a validated technical result into a commercial pilot. My roadmap is concrete. Fine-tune ESM-2 on the SKEMPI 3K mutation set to reach AUROC 0.70 or higher on that harder benchmark, then enter a pilot with Servier in Suresnes. The technical work is done. What remains is business development, partnership structuring, and go-to-market execution. That is the work a four-month incubator sprint is built for. My background is the differentiator. As a pharmacist, I have dispensed drugs that failed because resistance emerged in the clinic. As an ML engineer, I know how to build models that generalize. That combination is rare. I have seen resistance at the patient level, not just read about it in the literature. The venture targets oncology, antivirals, and antimicrobial resistance, three areas where resistance prediction directly changes treatment decisions. The programme is not biotech-specific, and that is acceptable. The open structure allows me to access Helsinki's network, test the venture against founders outside my field, and sharpen the pitch for later applications to EIC Accelerator and BPI i-Lab. I have already submitted to IncubAlliance and AI House in Paris. The Journeys Incubator Programme is the next validation step, not the final destination. I am committed to the full four months, on the ground in Helsinki, working daily on the venture. The technical risk is retired. The commercial risk is what I am here to retire next. RESEARCH STATEMENT The venture predicts drug resistance mutations from protein sequence alone, eliminating the dependency on crystal structures that limits current tools. Most resistance prediction methods require a resolved three-dimensional structure of the target protein. For many clinically relevant proteins, especially newly identified variants and proteins from emerging pathogens, that structure does not exist. My method removes that bottleneck. The technical foundation is a protein language model. ESM-2, trained on millions of natural protein sequences, learns evolutionary and biophysical constraints directly from sequence. I compute delta-embeddings, the difference between the embedding of a wild-type protein and its mutated variant. That delta captures the functional impact of the mutation in a high-dimensional space. I then concatenate the delta-embedding with an ECFP4 fingerprint of the drug molecule. A Random Forest classifier maps that combined representation to a binary resistance prediction. Validation results are specific. On the Platinum benchmark, 553 mutations with protein-grouped cross-validation, the model achieves AUROC 0.804 with a standard deviation of 0.025. This beats the published state of the art, mCSM-lig, which scores approximately 0.70. On SKEMPI 2.0, a harder benchmark with different mutation types, the model scores 0.634. That gap identifies the next technical milestone: fine-tuning ESM-2 on the SKEMPI 3K mutation set to push the SKEMPI AUROC to 0.70 or higher. Coverage is a second differentiator. Structure-limited tools can only score mutations in proteins with resolved structures, roughly 18 percent of the mutations in the Platinum benchmark. My sequence-based method scores 100 percent of them. In a clinical setting, that means a mutation observed in a patient sample can be assessed immediately, without waiting for a crystallization experiment or a homology model. The commercial application is a software platform for pharmaceutical companies. A pharma partner enters a target protein sequence and a candidate drug. The platform returns a resistance risk profile across all possible single-point mutations. That profile informs lead optimization, clinical trial design, and patient stratification. The first pilot target is Servier in Suresnes, with research collaborations at Paris-Saclay I2BC and Institut Pasteur, and a connection to Sanofi in Gentilly. The venture is pre-seed, proof-of-concept validated, and not yet incorporated. The next twelve months are defined: fine-tune on SKEMPI 3K, reach the AUROC target, sign the Servier pilot, and convert that pilot into recurring revenue. The Journeys Incubator Programme fits this timeline because it is a four-month intensive sprint, not a multi-year program. I can enter with a validated model and exit with a signed pilot agreement. SHORT ANSWER ESSAY Why this programme and why now The Journeys Incubator Programme is the right vehicle because my venture has passed the technical validation gate and now needs commercial acceleration. The model works. The benchmark numbers are public and reproducible. What I lack is the structured business development process that a four-month incubator provides: customer discovery with pharma partners, pricing model validation, and pitch refinement for later non-dilutive funding rounds. Helsinki's position as a growing biotech hub means the local network includes pharma, diagnostics, and computational biology players who can serve as early design partners. The timing is also driven by the roadmap. The Servier pilot is targeted within the next twelve months. A four-month intensive programme ending in late 2026 positions me to enter that pilot with a refined commercial package rather than a raw research prototype. What I will contribute to the cohort I bring a dual-domain skill set that is uncommon in most cohorts. I am a licensed pharmacist, so I understand the clinical reality of drug resistance, the prescribing context, and the failure modes of current therapies. I am also a machine learning engineer who has built and validated a model end to end, from data curation to benchmark evaluation. That combination means I can speak credibly with both the science team and the commercial team at a pharma partner. I also bring a working proof of concept with published benchmark numbers, which means the cohort gains a member with a validated technical asset, not just an idea. I am prepared to share my evaluation methodology with other founders working on AI applications in biology. How I will use the four months Month one: refine the customer discovery process with pharma partners, specifically testing the pricing and engagement model with the Servier pilot. Month two: complete the SKEMPI 3K fine-tuning and hit the AUROC 0.70 target. Month three: draft the pilot agreement structure and begin conversations with two additional pharma partners beyond Servier. Month four: prepare the EIC Accelerator application and the BPI i-Lab submission, using the incubator network for feedback on the proposal. The output is a signed pilot agreement and a submitted EIC Accelerator application. CHECKLIST - [ ] Confirm eligibility for Journeys Incubator Programme as a non-EU founder (verify visa requirements for Finland) - [ ] Verify the application deadline of 2026-08-09 and the exact submission portal requirements - [ ] Prepare a one-page technical appendix with the Platinum and SKEMPI benchmark results for the application - [ ] Draft a two-minute pitch video or slide deck if the programme requires one - [ ] Secure one letter of reference from a research collaborator at Paris-Saclay I2BC or Institut Pasteur - [ ] Confirm the Servier pilot timeline and obtain a written expression of interest if possible - [ ] Prepare a budget outline for the four-month programme, including relocation to Helsinki - [ ] Verify the legal status requirements: confirm whether incorporation is needed before or during the programme - [ ] Submit the application before the deadline with all required attachments EDITOR NOTES - Eligibility risk: the applicant is based in France and the programme is in Finland. Verify whether the Journeys Incubator Programme accepts non-Finnish founders and whether a residence permit or visa is required for the four-month period. This is the single biggest go/no-go risk. - The SKEMPI 2.0 AUROC of 0.634 is below the target of 0.70. The roadmap claims fine-tuning on SKEMPI 3K will close this gap. This is a stated target, not a validated result. Do not present it as achieved in any application material. - The Servier pilot is a named target, not a signed agreement. Confirm the current status of that conversation before submitting. If there is no written expression of interest, soften the language to "targeted pilot" rather than "pilot partner." - The venture is not yet incorporated. Some incubators require a legal entity for the programme. Verify whether incorporation is a condition of acceptance or can happen during the programme. - The profile lists multiple research lines. This application is specifically for the drug resistance prediction venture. Do not mix in other ventures or research threads. Keep the focus on the sequence-based resistance prediction platform.