Five practices, one accountable principal.

Practice Overview

We take raw sequencing, mass-spec or array data through QC, processing, statistics and interpretation. Every result comes with the methods, the code and the environment needed to reproduce it.

Core Capabilities

• Genomics

  Germline and somatic variant calling, structural variants, CNVs, annotation and prioritisation. Short-read and long-read (ONT, PacBio).

• Transcriptomics

  Bulk RNA-seq differential expression, splicing and fusion detection. Single-cell and spatial: clustering, annotation, trajectories, cell–cell communication.

• Epigenomics

  ATAC-seq, ChIP-seq and CUT&Tag peak calling, differential accessibility, motif analysis. Bisulfite and nanopore methylation.

• Proteomics & metabolomics

  Label-free and TMT quantification, imputation, differential abundance, pathway enrichment.

• Microbiome & metagenomics

  Taxonomic profiling, assembly and binning, functional annotation, AMR gene detection, diversity statistics.

• Multi-omics integration

  Factor models, network integration and joint embeddings that connect layers instead of stapling results together.

Key Deliverables

• ✓Analysis-ready matrices, variant tables and annotated objects
• ✓A written report with methods you can paste into a manuscript
• ✓Publication-quality static figures and interactive views
• ✓Reproducible notebooks with pinned environments

Practice Overview

A pipeline is software. We build yours with version control, tests, containers and CI, so it runs the same way for every sample, on every machine, and can be handed to the next person without a walkthrough.

Core Capabilities

• New pipelines

  Nextflow to nf-core standards, Snakemake or WDL, depending on your team and infrastructure. Modular, documented, parameterised.

• Legacy migration

  Turn a folder of Bash, R and Python scripts into a tested workflow without changing the science it encodes.

• Containers & environments

  Docker and Apptainer images, Conda and pixi lockfiles, and a strategy for keeping them current.

• Execution anywhere

  SLURM, PBS, LSF on-prem; AWS Batch, Google Batch and Azure in the cloud; Seqera Platform for orchestration.

• Testing & CI

  nf-test and pytest suites, small test datasets, GitHub Actions or GitLab CI that run on every change.

• Cost & performance

  Runtime and cost profiling, resource right-sizing, spot and retry strategies that cut cloud bills without losing runs.

Key Deliverables

• ✓A versioned pipeline repository with CI, tests and docs
• ✓Container images and resolved, reproducible environments
• ✓Runbooks for your cluster or cloud account
• ✓Runtime and cost benchmarks on your data

Practice Overview

Retrieval-augmented generation done properly: your documents, databases and live APIs behind an agent that plans, retrieves, checks its own work and cites what it used. Runs on your infrastructure or entirely offline.

Core Capabilities

• Ingestion

  PDFs, ELN and LIMS exports, internal reports, sequence and variant databases, wikis and tickets. Chunking and metadata that match how scientists search.

• Hybrid retrieval

  Dense embeddings plus BM25 with reranking, tuned on your own query logs. Domain embeddings for gene, variant and compound names.

• Agentic loops

  The agent decides when to query PubMed, Ensembl, UniProt, ClinVar or your SQL, and when to stop. Multi-step questions get multi-step plans.

• Grounded answers

  Every claim links to a source span. Unsupported claims are flagged, not smoothed over.

• Evaluation

  Retrieval recall, faithfulness and answer quality measured on a test set built with your team, before and after every change.

• Deployment

  Containerised service with an API, access control and audit logs. Cloud, on-prem or fully offline with an on-device model.

Key Deliverables

• ✓A retrieval service and API you own
• ✓An evaluation suite with baseline numbers
• ✓Ingestion pipelines that keep the index fresh
• ✓An admin interface for sources and permissions

Practice Overview

We build agents that operate inside your existing tools: they launch pipelines, read QC reports, curate literature, draft documents and hand off to a person at the points you choose. Every action is logged and reversible.

Core Capabilities

• Analysis operators

  Agents that launch Nextflow runs, watch for failures, retry sensibly and summarise results when a run completes.

• QC triage

  Read MultiQC and pipeline metrics, flag outlier samples, propose re-runs, and explain why in plain language.

• Curation assistants

  Literature screening, variant curation drafts, and structured extraction from papers into your schema.

• Report drafting

  Structured outputs into your templates: methods, results, figure captions, with links back to the data.

• Tool integration

  Model Context Protocol servers for your LIMS, ticketing, Slack, cluster and cloud, with scoped permissions.

• Safety & oversight

  Approval gates, budgets, rate limits, full trace logs, and evaluation and red-teaming before anything touches production.

Key Deliverables

• ✓Agents with defined tools, permissions and escalation paths
• ✓Trace logs and dashboards for every action taken
• ✓An evaluation suite and a staged rollout plan
• ✓Training for the team that will run them

Practice Overview

For narrow, well-defined tasks, a small model trained on the right data beats a large general one, and it runs on the MacBook already on your desk. We select, fine-tune, quantise and package models that work fully offline.

Core Capabilities

• Model selection

  Baseline evaluations across open model families (Qwen, Llama, Gemma, Phi and others) on your actual task, with licensing checked.

• Fine-tuning

  Supervised and preference tuning with LoRA and full fine-tunes; distillation from frontier models when you lack labelled data.

• Quantisation & packaging

  4-bit and 8-bit quantisation for MLX, llama.cpp and Core ML. Accuracy loss measured, not assumed.

• Native deployment

  Signed macOS apps, command-line tools or a local API server. Works on a plane, in a clean room, or behind an air gap.

• Profiling

  Tokens per second, memory footprint, thermal and battery behaviour on the M-series chips your team actually uses.

• Lifecycle

  An evaluation and update pipeline so the model improves as your data does, without regressions.

Key Deliverables

• ✓Model weights, the training recipe and an evaluation report
• ✓A signed macOS app or CLI, or a local server
• ✓Fine-tuning and evaluation code you own
• ✓An update playbook for the next version