ML Engineer, Discovery Applications

Mithrl•San Francisco, CA

1d•Onsite

About The Position

We are hiring an ML Engineer, Discovery Applications to build the high level, end-to-end scientific workflows that power real bench to bench decision making inside the Mithrl platform. This role focuses on building the application layer on top of the AI Co-Scientist. Your work will shape how scientists discover biomarkers, identify and validate targets, design experiments, and run early discovery programs that extend all the way to IND-enabling work. This role requires a deep understanding of the discovery and preclinical development cycle. You should understand how research teams move from early target hypotheses to biomarker strategy, hit identification, hit to lead, lead optimization, and preclinical validation. Your applications will support decision making across this entire arc and will be consumed directly by scientists and program teams. You will design multi step workflows that combine analysis modules, ML models, domain logic, and agentic reasoning into complete applications. These applications cover biomarker discovery, target ID, target validation, small molecule hit identification and optimization, and gene therapy workflows. You will also extend applications to support new data modalities as our platform expands.

Requirements

Strong experience in ML, computational biology, scientific computing, or a related field
Deep understanding of the drug discovery and preclinical development cycle including early discovery, target identification, target validation, hit identification, hit to lead, lead optimization, and IND-enabling work
Experience building analytical workflows or application logic for biological or scientific data
Familiarity with key discovery analysis methods such as differential expression, pathway analysis, clustering, enrichment, and target scoring
Proficiency in Python and scientific computing libraries and comfort with building multi step workflows
Ability to convert scientific questions into structured, reproducible workflows that support real decision making
Strong communication skills and ability to collaborate with cross functional engineering and biology teams

Nice To Haves

Experience building LLM powered agents or multi agent reasoning systems
Experience with multi modal biological data integration
Experience with computational chemistry tools such as docking or ADMET modeling
Familiarity with biological ontologies, curated knowledge sources, or pathway databases
Prior experience in a tech bio startup, biotech R&D group, or scientific software platform

Responsibilities

Build full discovery applications that support biomarker identification, target discovery, target validation, small molecule design workflows, and gene therapy programs
Stand up new analyses that support application logic and improve or extend the existing analysis suite
Create multi step reasoning flows that integrate ML models, statistical methods, pathway context, simulation tools, and biological domain logic
Design application specific workflows for compound evaluation, program prioritization, and multi modal evidence integration
Extend existing applications to incorporate new data modalities and new analysis routines
Build reusable frameworks for Design of Experiments across biomarker discovery, target ID, validation, small molecule programs, and gene therapy
Implement and improve the AI systems that orchestrate and chain analyses into coherent applications used directly by scientists
Collaborate closely with ML engineers, bioinformatics teams, and data ingestion teams to ensure workflows run on consistent data
Validate scientific correctness and ensure applications produce accurate, reproducible, and interpretable results