Axle Informatics

• Architect, develop, and maintain AI-powered applications in Python, including conversational assistants, agentic tools, and retrieval-augmented generation (RAG) systems. • Design and optimize retrieval pipelines: embedding generation, vector search, hybrid search, and reranking against large domain-specific corpora. • Build document ingestion and parsing pipelines that extract structured content (text, tables, figures, citations) from scientific and technical documents for downstream use by AI systems. • Integrate structured knowledge — ontologies, controlled vocabularies, and knowledge graphs — with unstructured text to produce evidence-grounded, provenance-aware AI outputs. • Develop and operate shared platform services that accelerate delivery of new AI applications: model gateways, agent-tool integrations, authentication, API key management, and observability. • Lead technical design decisions and establish best practices for code quality, testing, evaluation, and deployment of AI systems. • Provision, deploy, monitor, and troubleshoot services on Kubernetes and cloud infrastructure, including container builds, manifests, and CI/CD pipelines. • Design and maintain relational and vector databases, including schema design, indexing, migrations, and incremental data updates. • Define and track quality metrics for AI systems (retrieval quality, answer faithfulness, latency, cost) and drive measurable improvements over time. • Engage directly with non-technical clients to gather requirements, clarify goals, manage expectations, and validate that delivered solutions align with their objectives. • Translate high-level client and mission needs into actionable technical specifications and development plans. • Mentor junior engineers and contribute to a culture of technical excellence. • Participate in and help facilitate standups, sprint planning, and cross-team review meetings across multiple project streams.

• Develop, implement, and optimize LLMs and computer vision algorithms for processing and analyzing biomedical text and image data. • Design and develop software tools and pipelines for automated image analysis, data integration, and visualization, leveraging machine learning and deep learning techniques. • Collaborate with biologists and other researchers to understand their needs and translate them into technical solutions, incorporating state-of-the-art AI methodologies. • Perform data mining, statistical analysis, and machine learning on large-scale biomedical imaging datasets to extract meaningful insights and improve predictive accuracy. • Validate and benchmark AI-driven imaging algorithms and tools to ensure their accuracy, robustness, and scalability. • Document software development processes, create user manuals, and provide training and support to end-users, particularly focusing on AI applications. • Stay up-to-date with the latest advancements in AI, computer vision, biomedical imaging, computational biology, and data science, and apply this knowledge to improve existing tools and methodologies. • Participate in interdisciplinary team meetings, present findings, and contribute to scientific publications and grant proposals, emphasizing AI and computer vision innovations. • Identify and implement new AI technologies and frameworks to enhance the capabilities of biomedical imaging tools and methodologies. • Knowledge of code packaging, containerization technologies, and continuous integration pipelines. • DevOps experience with Kubernetes and workflow automation is a plus.

• Evaluate existing bioinformatics pipelines for performance, accuracy, and maintainability, identifying opportunities for optimization and enhancement • Develop, extend, and maintain Nextflow workflows for bulk RNA-seq, single-cell RNA-seq, DNA variant calling, methylation analysis, and emerging assay types • Architect scalable solutions capable of processing thousands of samples efficiently on Google Cloud Platform • Optimize data input/output operations across pipeline modules to minimize bottlenecks and reduce computational costs • Prepare and structure analysis outputs for visualization platforms and downstream statistical or machine learning applications • Implement best practices for workflow versioning, containerization, testing, and documentation • Collaborate with computational biologists, data scientists, and software engineers to integrate pipelines into broader analytical ecosystems • Stay current with advances in sequencing technologies and analytical methods, evaluating and incorporating new tools as appropriate

• Design, develop, and maintain high‑performance C++ software for scientific and computational workloads. • Build scalable applications using modern C++ (17/20) with an emphasis on correctness, performance, and maintainability. • Develop and extend scientific computing pipelines running in Linux and HPC environments. • Optimize compute‑intensive code using profiling, benchmarking, and performance‑tuning techniques. • Collaborate with senior engineers and researchers to translate computational requirements into robust software solutions. • Work with containerized applications (Docker, Singularity) deployed in HPC and cloud environments. • Integrate applications with HPC schedulers (e.g., Slurm) for batch and interactive workloads. • Contribute to documentation, testing, and CI/CD workflows to support reliable software delivery.

• Build production-grade data systems by writing clean, modular, well-tested Python code for data pipelines and platform services. • Design and maintain data models aligned with medallion architectures (bronze/silver/gold). • Build and maintain data workflows using Dagster (preferred) or Airflow. • Implement idempotent transformation logic using SQLMesh/Tobiko (preferred) or dbt. • Build, debug, and optimize Docker images for local and production environments. • Read and modify infrastructure-as-code using Terraform. • Document data flows, system architecture, and operational procedures clearly.

• Nextflow Refactoring & Architecture: Identify imaging pipeline components suitable for direct implementation in Nextflow and refactor them into modular, reproducible processes that mirror or improve upon validated behavior. • Deep Imaging Data Analysis: Define and document required inputs, intermediate outputs (segmentation masks, feature tables), and final plate-level summaries, ensuring the engineering supports the scientific intent. • Architect for Heterogeneous Hardware: Design Nextflow configuration profiles that allow identical workflow definitions to run across diverse environments, from high-end GPU clusters (Nvidia L40s) with SLURM to legacy on-premise HPC nodes. • Headless Integration: Lead the development of scripts and wrappers that enable headless execution of imaging logic, integrated directly into a GCP Workflows-based automated control system. • Scientific Validation: Design and execute regression checks to ensure that refactored Nextflow pipelines produce outputs equivalent to legacy systems within strict scientific tolerances. • Modular Pipeline Design: Ensure every step in the pipeline is as stateless and modular as possible to allow for rapid refactoring and reuse across different imaging assays. • Integrate the “Lab Loop”: Collaborate with engineers to manage data handoffs between on-premise edge infrastructure and cloud-based system-of-record databases (PostgreSQL/ARC).