Senior Software Engineer, Genomics

Title: Senior Software Engineer, Genomics Location: Remote

About The Role

Feanix’s main product Autopilot is a farm decision system. It tells dairy farmers what to do with every animal in the herd: which cows to breed and with what semen, which to cull, which to sell, and which heifers are worth raising. Those decisions are driven by herd records, lactation models, farm economics, and genomics, with one aim: making each farm as efficient and profitable as possible.

Genomics at Feanix runs from sample to result. We collect hair samples from animals, run our own sequencing process, and impute full genotypes against a reference panel. Those genotypes tell us what each animal is really worth, confirm who its parents are, and help farms avoid inbreeding as they choose which bulls to breed to which cows.

This is a production role, not a research one. Our genomics work supports research, but above all it feeds the decisions that run working dairies. Genotypes don’t make those calls alone: combined with each herd’s phenotypic data (milk, reproduction, health, and lifetime records), they decide which animals to breed, keep, and sell, every week.

The base methods behind these pipelines are strong, and so is the pipeline we’ve built to run it. What comes next is expanding it for much higher volume, tighter accuracy, and lower cost per sample as we grow from 20+ farms to hundreds. And dairy is where we started, not where we stop: the pipeline you build should be ready to carry other species and genomics uses beyond the herd decisions we deliver today.

We're building out genomics as a core capability of engineering, and this is our second genomics hire. You'd partner with our founding genomics engineer, who brings deep knowledge of the genetics and the methods behind them. You bring the production engineering. You’d lead how our genomics pipelines are built and run, taking them from manually orchestrated to fully automated under code, CI, and monitoring, and set the engineering standards for genomics as the team grows. Their domain depth plus your production experience is what makes the science run unattended, observably, and repeatably, every week, at hundreds of farms.

What You'll Own

The genomics team owns everything below together; you’re the one who leads on making it production-grade. The work is continuous: keep making the genomics processing pipeline more automated, more reliable, faster, and cheaper per sample, so that adding a farm adds samples, not engineering work.

• The genomics pipeline as a production system. Sample → sequence → align → impute → genotype → parentage/trait, scheduled or triggered by new data, with retries, version pinning, and alerting.
• A sample-state ledger. One authoritative record per sample of where it is (received, sequenced, passed coverage, imputed, called) and, if it stopped, why.
• Parentage, end to end. Our parentage results already reach farms, built on the same genotypes, sample identities, and compute as the rest of genomics, so the team owns it end to end, from sample to result. You’d move fleet runs onto a schedule with completion and lateness alerts, and finish the readiness work that makes our already-built successor runner the system of record.
• CI and deploys, in partnership with infrastructure. Our infrastructure engineer designs the CI and deploy path across engineering. You’d work with them to bring the genomics repositories onto it, so every change is tested and every pipeline and workflow version is managed in code and reviewed.
• Cost and storage. A cost-per-sample view for sequencing compute and storage, and a lifecycle policy for data measured in hundreds of terabytes.
• Lab and vendor intake. The handoff from our in-house sequencer, outside sequencing labs, and genotyping-chip providers into the pipeline.
• Reproducible model training. Making breeding-value model training runnable and repeatable, so time goes to methods, not reruns.

You’ll work closely with others who own the pieces on either side. Our founding genomics engineer owns methods and confidence thresholds. You should understand them well enough to debug and push back, not redesign them. Model selection and promotion into breeding decisions, delivery of results into farmers’ herd software, and the platform-wide animal identity key each sit with other engineers. The genomics team owns sample identity and how samples attach to that key.

How We Work

Python end to end, with uv and pytest, on AWS. Genomics runs on a mix of compute chosen per workload: Modal for serverless Python, Nextflow workflows on AWS HealthOmics, AWS Batch, and GPU-accelerated alignment. Data lives in S3. The genomics toolchain includes Parabricks, samtools, bcftools, PLINK, GLIMPSE2, Beagle, KING, and JWAS. Terraform/OpenTofu and GitHub Actions provision and ship it. This is what you’ll work with, not a checklist to match.

Like everyone else, we're working out how humans and agents build software together. Our engineers direct agents against real work and build the harnesses that check that work. Quality comes from what happens around the code: specs, tests, verification. Nobody has this fully figured out, and that includes us.

What You Bring

• Several years running scientific or bioinformatics pipelines in production (Nextflow, AWS HealthOmics or Batch, or similar), scheduled, monitored, and depended on by people other than you
• Enough fluency in genomic data to debug it: FASTQ/BAM/VCF, PLINK, imputation (GLIMPSE, Beagle, or similar), and coverage and QC metrics
• Experience building sample tracking, LIMS-like state, or another “where is every item, and why” system
• Strong Python, and comfort with S3-scale data, idempotent pipelines, retries, and alerting
• Cost awareness with large data: storage lifecycle and compute spend per unit
• A track record of working alongside scientists, translating methods into reliable systems and asking good questions about them
• Daily, load-bearing use of coding agents, and a real position on where they break down

Preferred Qualifications

You've likely held a title like bioinformatics engineer, data engineer, or software engineer on a genomics or research-platform team. That title matters less to us than two other things: that you think creatively about how agents are changing this job, and that you're motivated by taking a working pipeline and making it run at 10x the volume, cheaper and more reliably.

• Livestock, plant, or human population genomics
• Parentage or pedigree work: SNP-based parentage verification, kinship, or reconciling recorded pedigrees with genotypes
• Production experience with genotype imputation or sequencing pipelines (alignment, variant calling)
• Genomic prediction (GBLUP and relatives); useful, not expected
• LIMS or sample-tracking integration with a wet lab
• Having supported a small team through a period of rapid growth
• Time at a sequencing lab or genomics service company
• Terraform/OpenTofu
• Agriculture experience isn't required, but an interest doesn't hurt

Ultimately, we care far more about demonstrated ability, curiosity, and problem-solving than pedigree or conventional career paths.

Logistics

• Reports to the Platform Engineering Manager
• Remote (US), with 10am–2pm Pacific core hours regardless of where you're based
• Travel 15–20%, including visits to commercial dairies and time with the team at our headquarters in Scotts Valley, CA
• Engineers share an on-call and triage rotation: weekly, with a written handbook, and weekend work requires explicit sign-off rather than being assumed. Building the alerting that keeps genomics on-call quiet is part of this job.

Why You'll Enjoy Working With Us

• The People: we have an amazing group of talented, smart and fun people on our team!
• The Company: we are building something novel that has not been done before which is making a huge impact on the dairy industry!
• The Experience: you get to come in and be a part of ground breaking work that will have a huge impact on generations to come!