Sr. Scientist I/Sr. Scientist II, Applied AI & Agentic Systems for Drug Discovery

Antares Therapeutics•Boston, MA

1d•$160,000 - $220,000•Hybrid

About The Position

We are seeking a scientist who builds. As a member of a highly integrated team of medicinal chemists, chemical biologists, computational scientists, biologists, and machine learning scientists, you will develop and deploy LLM-driven agents for scientific workflows that shorten our design-make-test-analyze cycles and sharpen program decisions. You will sit with project teams and understand the questions that actually gate progress: which compounds to make next, whether a target is tractable, what the assay and DMPK data are telling us. You will turn them into agentic systems that reliably orchestrate our tools and models to maximize the usage of our data. In practice that means software that carries out multi-step work on a scientist’s behalf: retrieving the relevant data, running analyses, checking outputs, and returning sourced answers. Multi-agent systems, Model Context Protocol (MCP) servers, and reusable skills for LLM-based tools are the means; better and faster discovery decisions are the end. This is a hands-on individual-contributor role on the Antares scientific track, for someone who has already contributed to drug discovery programs and wants to multiply what a lean discovery organization can do. The role is offered at the Senior Scientist I or Senior Scientist II level; scope, independence, and cross-functional influence scale with experience, as described under Qualifications.

Requirements

PhD. in computational chemistry or biology, cheminformatics, computer science, machine learning, or a related quantitative field (or equivalent experience).
Senior Scientist I: PhD with 5+ years (or bachelor's with 13+ years) of relevant experience.
Senior Scientist II: PhD with 8+ years (or bachelor's with 16+ years), with a demonstrated record of independent delivery and cross-functional influence.
Demonstrated, direct contributions to small-molecule drug discovery programs in a pharmaceutical or biotechnology setting — where your work measurably influenced design decisions or program progression. This is central to the role.
Hands-on development and deployment of LLM-driven agents and multi-agent systems for scientific workflows, including MCP servers and tool integrations, using current agent frameworks and model-provider SDKs with frontier models (e.g., Claude, GPT). We are framework-agnostic: what you build matters more than what you built it with.
Strong programming in Python, with fluent, effective use of coding agents such as Claude Code or Codex.
Experience with model evaluation and benchmarking, and with MLOps practice for reliable, reproducible deployment.
Working knowledge of medicinal chemistry, SAR, structural biology, and DMPK principles sufficient to be a credible partner to drug hunters; familiarity with cheminformatics toolkits (e.g., RDKit, OpenEye, Schrödinger) is strongly preferred.
Excellent written and verbal communication skills, with a proven ability to drive adoption among scientific colleagues and to make complex technology approachable.
A passionate, results-oriented drug hunter who thrives in a collaborative, data-driven, and scientifically ambitious culture, and who measures success by program impact rather than tools shipped.

Nice To Haves

Familiarity with cheminformatics toolkits (e.g., RDKit, OpenEye, Schrödinger)

Responsibilities

Embed with discovery project teams as the AI enablement partner, and work alongside medicinal and computational chemists, chemical biologists, and machine learning scientists to identify where agents can accelerate design-make-test-analyze cycles and influence program decisions.
Build and maintain MCP servers and tool integrations that connect our LIMS/ELN, structural, assay, and DMPK data to agentic workflows, respecting existing data permissions, so scientists can interrogate our own data directly.
Develop and deploy LLM-driven agents and multi-agent systems for scientific workflows, orchestrating tools and models to, for example, enable assay data analysis, SAR interpretation, target assessments, DMPK understanding, and literature and competitive intelligence.
Own evaluation and benchmarking of what you deploy, supported by sound MLOps practices — defining task-specific benchmarks, running error analysis, and tracking accuracy and reliability so outputs are trustworthy enough to inform go / no-go decisions.
Develop reusable skills for LLM-based tools and reference implementations, so a solution built for one program serves the whole portfolio.
Analyze and integrate data across functions to independently identify, define, and address critical scientific questions; communicate approaches, limitations, and recommendations clearly to project teams and leadership, and help colleagues use these tools confidently.
Act as the AI enablement function lead on programs, establish internal best practice, and evaluate and recommend platforms and emerging methods.