Senior Automation Engineer

Automation EngineerSatomic's mission is to close the gap from idea to molecule with faster navigation of chemical space. We are building a robotic chemistry platform that integrates laboratory robotics, software, and AI to transform small-molecule synthesis and drug discovery.Satomic has raised a $15M Seed from fantastic investors including Riot Ventures, HOF Capital, and Compound VC, and our team includes alumni from leading large pharmas, small biotechs, and AI labs.You will build the automation and data tooling our synthetic chemists use to develop and validate reactions, and you will own process quality in our data generation pipeline. On the build side, that means taking an experimental design and turning it into something the automation can actually run: plate maps, worklists, and the parameters the instruments need, generated rather than assembled by hand. On the quality side, it means the controls and acceptance criteria that hold execution consistent plate to plate and month to month, and the experiments that find where quality is slipping so we can keep raising it.This role is critical to Satomic's mission of closing the gap from idea to molecule by making sure that every experiment we spend instrument time, material, and people on is well-designed to answer the question behind it.You will own:The automation tooling our synthetic chemists work with. Building it, refining it as they use it, and closing the gap between what a chemist wants to run and what the instruments can execute.What the synthetic chemistry team hands to data generation. You set what a condition set has to prove before it runs at scale: plate design, controls, replication, and what counts as validated. The chemist owns the reaction and the conditions; you own whether the experiment was built so the results hold up in production.Process QC in the data generation pipeline. The controls, anchors, replication, and plate acceptance criteria that catch a bad run while it is still a bad run, instead of weeks later when a model underperforms and nobody can say why.Execution quality once campaigns are running. Whether a plate was run to standard, whether results are consistent across operators and decks and months, and what happens when a run fails partway through.Finding where the process is losing us quality and fixing it. The side experiments that separate a real result from an artifact, pin down sources of bias as new reactions come in, and keep moving the standard up rather than holding it steady.Engineering new capabilities in the reaction process that unlocked experiments we couldn't run before, to extend the platform's reach and continuously make the data more reproducible.By day 30, you will have:Mapped reaction development through the handoff into data generation, and identified the friction points worth removing: the manual steps, the rework, and the places information gets lost.Shipped one method or tool that a chemist uses in their real work. What matters is that it went into use and that you learned where the friction actually is by building against it rather than asking about it.Established a baseline metric for data quality and reproducibility on current campaigns, so every improvement after this is measured against a real number rather than asserted.By 90 days, you will have:Created a workflow where every step of reaction development is supported by automation a chemist can rely on, rather than a mix of manual work and scripts that only their author can run.Identified three opportunities to improve process quality and experimental results, and built them into the data generation pipeline rather than leaving them as recommendations.Taken two reactions from development through to data generation campaigns that run end to end, run to the standard you set: controls in place, plate acceptance called on evidence, and results usable by ML.By year 1, you will have:Created a pipeline where onboarding a reaction and running it at scale is a repeatable path rather than a bespoke effort each time. Chemists, data generation, and ML all know what they are handing off and what they are receiving and the campaign runs to a known standard.Made the parameterization and design tools the path chemists reach for by default, and began encoding that judgment into agent-executable process, so the standard scales without you in the room.Reduced campaign-level noise or material waste on a named metric, so we win more information per well and get the same answer next month.By year 5, you will have:Made screening discipline native to how the company works, carried by a cohort of chemists and data-generation scientists who design this way by default and teach it onward, so the standard no longer depends on you and holds when you move to the next hard problem.Encoded your design judgment into agents that critique and correct campaign designs before they run without a human in the loop so rigor is enforced on every experiment.Made hard reactions scalable by solving the development problems creatively, then engineering the solution into something that runs at scale reproducibly.This role is:Uncomfortably fast-moving. Planning ahead down the question-tree to prioritize follow-on experiments and react in real-time as data comes in is critical.A hybrid scientist and engineer role. It takes real statistical rigor and real time at the bench and the instrument.A role for someone who cannot leave a bad process alone, and who is expected to look beyond their own scope. If something is broken two teams over, go find out why.Embedded in Automation, but whose time is spent mostly with domain experts who are neither screening nor automation people. Meeting synthetic chemists and data scientists where they are is most of the job.Teaching and influence. You raise the standard of work you may not individually own.This role is NOT:A role where building is the whole job. You will write code and build methods (plenty of them), but you are accountable for whether what you build produces data we can trust, not just whether it runs.A role where you only guide other automation engineers. Much of your influence lands on chemists and other colleagues who have never worked in a high-throughput setting.A role with a clear standard or playbook to execute. We expect you to be opinionated, learn quickly, and drive change within our team.Qualifications:Hands-on high-throughput experimentation. You have scaled up a process from low-throughput to high-throughput (both design and execution), and set the process checks and acceptance criteria.Design of experiments fluency: factorial and fractional designs, blocking, randomization, replication and power, and splitting observed variance into its sources.You have chased down a problem in a physical process and proved what caused it, rather than settling for a plausible story.Teaching. You have brought people without your background up to your standard and left patterns behind rather than advice. You have also watched your own protocol run in someone else's hands and changed the design because of what you saw. You can codify your intuition and scale it with SOPs or non-human agents.You write your own scripts to handle data. Reformatting instrument exports, cleaning up spreadsheets, getting records from one system into another: you automate the tedious parts instead of clicking through them.AI literacy and the habit of using AI tooling to extend your own reach, including an interest in turning your judgment into an agent-executable process. Python and enough data skill to interrogate your own campaign data without waiting on anyone.Enough scientific background to get up to speed quickly on synthetic chemistry to talk to a chemist as a peer.Overview of Satomic's Interview Process:Phone screen (15 minutes): A brief introductory call to discuss the role and answer questions.Technical screen (30 minutes): A discussion with a technical colleague on your experimental design depth and screening experience.Take-home assignment (4-6 hours): A scoped experimental design problem representative of the work at Satomic, completed on your own time.Final round (75 minutes): A presentation of a screening or design campaign you have owned, then a case discussion and conversations with Satomic's leadership team on how you work, how you teach, and how you collaborate across disciplines.Compensation:This role offers a base salary of $1