Senior Bioinformatician / Data Scientist with a PhD in Computational Biology (2022) and 8+ years building production ML pipelines, statistical models, and cloud workflows from raw sequencing reads to decision-ready reports. Currently Lead Bioinformatician at Genics. Lead data scientist on $3.2M+ NIH / UNDP / DFAT-funded research at CSIRO + Mayo Clinic. Quantitative analyst at AE Capital Hedge Fund. Builds agentic AI systems on the side.
I work in the messy middle between research-grade methods and production infrastructure: the place where raw sequencing data, statistical assumptions, cloud systems, stakeholder needs, and deadlines all collide.
At Genics I lead bioinformatics strategy across commercial genomics: end-to-end Nanopore 16S pipelines for microbiome and pathogen detection, anomaly detection on noisy signals, automated client reporting, genomic selection databases, and the SOPs that keep all of it reproducible.
Before that I spent three years at CSIRO co-leading cloud genomics and personalised medicine programs with Mayo Clinic, Université Paris-Saclay, and partners in Indonesia and Korea. In charge of the FHIR + blockchain consent infrastructure (GeneGuardian), and built pharmacogenomics and variant detection workflows (PharmCAT, sBeacon, VariantSpark, DeepVariant) alongside pangenome graph methods. Lead data scientist on $3.2M+ NIH, UNDP and DFAT-funded research.
Quantitative finance is something I genuinely love. I had a first taste at AE Capital Hedge Fund validating ML signals on multi-asset time-series, and I would love another shot at it. In the meantime I keep that muscle alive through my agentic AI work, where one of the live use cases is algorithmic trading research. The pattern across genomics, pathogens, and markets is the same: high-dimensional noisy data, statistical rigour, and decisions that have to hold up under scrutiny.
On the side I build agentic AI infrastructure (OpenClaw + Hermes), orchestrating Claude, GPT, and local Ollama models to automate research, document workflows, and business operations.
Architected, set up and deployed the company's full Oxford Nanopore analytical stack: live DNA sample to pathogen detection report. Covered basecalling, mapping, variant calling, pathway analysis, statistical testing, SQL integration, and automated client reporting. Validated and reliability-tested end to end.
Co-led an AI-assisted dynamic consent platform for genomic data: FHIR integration, blockchain, decentralized storage (IPFS / FileCoin / Storj), AWS + Terraform deployment. HIPAA / GDPR / GA4GH compliant. Now live and serving customers. Supervised 10+ students across IT, medical informatics, and bioinformatics.
Deployed and trained partners on end-to-end pipelines hosted in their own AWS accounts across Indonesia, Korea, Mayo Clinic, and Australian clinical sites, supporting pharmacogenomics, infectious disease surveillance (COVID-19, TB), and data sovereignty. Stack covered PharmCAT (pharmacogenomics), sVEP (variant effect prediction), VariantSpark (random-forest GWAS), and sBeacon (federated genomic data exchange).
Validated machine-learning signals and built statistical models on high-dimensional financial time-series across equity prices, bond valuations, and macro indicators. Feature engineering, anomaly detection, EDA, and risk management in a fast-paced hedge-fund environment.
Sole developer for a commercial-grade SNP array pipeline on a new tree species, leading the work from scoping and resource/price estimation through validation, delivery, and client-owned IP handover. Also built the relational SQL database (DbSchema) for downstream genomic selection.
Open-source multi-agent orchestration framework routing tasks across Claude API (reasoning + code), OpenAI GPT (general-purpose), and local Ollama Qwen models (lightweight heartbeat agents). Live use cases: algorithmic trading research, personal AI assistant, automated webpage generation, business-development automation, plus research and bibliography workflows driven by genuine curiosity.
Led a multi-institution collaboration with Mayo Clinic and Université Paris-Saclay applying pangenome graph methods to improve precision medicine for Somali and other underrepresented Horn of Africa / Middle Eastern populations. Directed 10+ contributors across IT, bioinformatics, and biostatistics. Quantified reference bias effects on PCA, D, and F-statistics; ran the full biostatistics analysis on HPC.
Benchmarked ancient-DNA alignment methods, quantified reference bias effects on PCA, D, and F-statistics, and built pangenome graph workflows that materially improved analyses for underrepresented populations. Trained Google DeepVariant (deep-learning variant caller) on aDNA-specific damage patterns.
Leading bioinformatics strategy and end-to-end pipeline development for a commercial genomics company serving agricultural and industrial clients.
Led and delivered 4+ concurrent national and international programs spanning cloud genomics, AI-assisted bioinformatics, and federated learning.
Validated ML signals and modelled high-dimensional financial time-series across equities, bonds, and macro indicators in a fast-paced hedge-fund environment.
Thesis: "Quantifying and Reducing Biases in Genomic Research Using Pangenome." Three Q1 first-author papers, co-authored FAIR/CARE community standards for reproducible aDNA research.
Built OpenClaw + Hermes, a system that autonomously designs and coordinates workflows across multiple LLMs. Claude API handles complex reasoning and code generation, OpenAI GPT models cover general-purpose tasks, and local Ollama-hosted Qwen models run as lightweight heartbeat agents. Each model specialises in what it does best, inside one coordinated environment.
Quantitative finance has been a long-standing passion. I had a first taste at AE Capital Hedge Fund and would love another shot at it. In the meantime I keep that muscle alive by merging it with my agentic AI work: algorithmic trading pipelines, signal validation, and anomaly detection on market data are some of the live use cases of the OpenClaw framework.
Outside work, I like staying active and curious: volunteering as a basketball coach, paragliding when I can, building casual coding prototypes, playing video games, and getting around a good board game. It is less about polishing a side-hustle brand and more about keeping learning loops alive.
Best fits: senior hands-on or tech-lead roles in bioinformatics, computational biology, or data science. Also genuinely keen on quantitative analyst or quant research positions. Open to Australia, Europe, or remote. Response within 48 hours.
adrioliva@hotmail.fr →