# Insilico Search Partners — Principal Agentic Scientist

- Generated: 2026-09-22 05:17:29 PM EDT
- Original/canonical posting: https://www.linkedin.com/jobs/view/4469017404/
- Provider: LinkedIn
- Posted: approximately 2026-09-22 12:17 PM EDT (from '5 hours ago')
- Elapsed since posting: represented by the provider's relative posting label and approximate Eastern timestamp when available
- Applicants: Be among the first 25 applicants
- Work model/location: Remote-hybrid — Cambridge, Massachusetts; monthly travel
- Compensation: $208,000-$275,000 annually
- Travel: Once a month; exact percentage and overnight requirement not disclosed
- Positioning track: Technical manager
- Fit outcome: PASS — 85%

## Direct-match strengths

Agentic AI architecture, Python, LangGraph, production LLM systems, evaluation and benchmarking, observability and cost governance, computational biology, genomics, literature mining, technical program leadership, patents, Ph.D.-level life sciences, and team mentoring.

## Hard or material gaps

Recent direct work in protein design, docking, cheminformatics/ADMET, multi-omics, and translational/clinical development is limited. Keith's strongest life-science evidence is older genomics, bioinformatics, pathway analysis, pharmaceutical research, and drug-target knowledge discovery, combined with current agentic-AI depth.

## Evidence map

1. Agentic AI/ML architecture (weight 3, evidence 3/3) — Direct agents, LangGraph, orchestration, evaluation, and production ownership.
2. Computational biology and genomics (weight 3, evidence 3/3) — Ph.D., genomics, pathway analysis, and bioinformatics products.
3. Protein design/docking/ADMET (weight 3, evidence 1/3) — Adjacent life-science depth; these specialties are not directly supported.
4. Python and modern ML frameworks (weight 3, evidence 3/3) — Direct Python, TensorFlow, SageMaker, and production AI.
5. Evaluation, observability, and cost (weight 2, evidence 3/3) — Direct evals, monitoring, routing, and cost controls.
6. Program and team leadership (weight 2, evidence 3/3) — Built technical teams and led complex programs.
7. Patents and high-impact delivery (weight 2, evidence 3/3) — Named inventor with globally adopted commercial bioinformatics work.
8. Location, travel, compensation (weight 2, evidence 3/3) — Massachusetts hybrid and pay qualify; monthly travel remains a caveat.

## Keyword diagnostic

Distinctive combination of current agentic-AI systems and earlier commercial bioinformatics/pharma evidence; specialized modern drug-design depth remains the main caveat.

## Full normalized job description

Principal Scientist, Agentic AI/ML Systems — Drug Discovery
Machine Learning | Remote-Hybrid, once a month travel | Cambridge, MA
$208,000 – $275,000
About the Opportunity
A fast-growing, AI-first, venture organization focused on accelerating pharmaceutical R&D is seeking a Principal Scientist to architect and lead agentic AI/ML systems across the drug discovery space, from target ID through translational and clinical development. This is a chance to build the foundation of LLM-based agents and agentic workflows, that ties genomics, protein design, multi-omics, cheminformatics, and literature mining into unified, end-to-end pipelines scientists rely on daily.
The Role
Lead the design, orchestration, and scaling of agentic AI systems that automate and accelerate computational biology workflows. Partner with venture and platform leadership to define pragmatic AI strategy, own technical execution, and set the benchmarking and evaluation bar for agentic systems.
What You'll Own
Program Leadership:
Lead multiple AI/ML and computational programs spanning preclinical, translational, and clinical R&D.
Agentic System Architecture:
Design, build, and scale agentic systems (LangGraph, CrewAI, AutoGen, Pydantic AI) that orchestrate ML and comp-bio tools — genomics, biomolecule design, docking, multi-omics, literature mining — into automated end-to-end scientific workflows.
Technical Ownership:
Own build, benchmarking, evaluation, and maintenance of agentic pipelines, including sandboxed code execution, agent harness development, and observability/cost-governance tooling.
Team Leadership:
Manage and mentor scientists/engineers; support recruiting and interviewing as the ML function grows.
Strategy & Landscape:
Track emerging agentic-AI/ML literature; translate it into build strategies that accelerate R&D.
Communication:
Translate complex technical work for cross-functional and executive audiences to drive decisions.
What You Bring
MS or PhD in Machine Learning, Statistics, Computational Biology, or a related field, with 5+ years applying AI/ML in academic, pharma, or biotech settings.
Demonstrated ability to lead agentic AI/ML projects end-to-end, from architecture to production.
Strong Python and modern ML framework depth (PyTorch or JAX/TensorFlow).
Track record of publications, patents, or high-impact technical delivery.
Preferred:
hands-on integration of LLM platforms (Anthropic, OpenAI, Vertex AI, Bedrock) and evaluation/feedback-loop frameworks for agentic systems; breadth across genomics, protein design, proteomics, cheminformatics/ADMET, or biophysics.

## Artifact metadata

- Resume: https://bit.ly/4hmgfUG
- Cover letter: https://bit.ly/4ji9GoO
- Validation: PASS — 2-page resume (930 words), 1-page cover letter (196 words); PDF geometry, bounds, annotations, links, and visual pages verified.
- Google Drive used: No