Manual Workflows Could Not Keep Pace with Design Request Volume
The client sells assay products used in genomics research, molecular analysis, cancer studies, pharmacogenomics, and agrigenomics. Researchers order custom assay designs tailored to the targets they want to study. The company then generates probe sequences, validates them against reference databases, and delivers design files that go to manufacturing.
Before this platform existed, that process was a tangle of spreadsheets, email threads, and manual handoffs. Scientists sent requests to internal teams who ran pipelines by hand, checked results, and uploaded files to shared drives. Turnaround was slow. Errors crept in. There was no single place to see the status of a request or trace the provenance of a design file back to the inputs that created it.
As order volume grew, the manual approach stopped scaling. The company needed a self-service system where scientists could submit requests, track progress, and receive validated outputs without anyone babysitting each job.
End to End Engineering for a Multi-Product Genomics Platform
Sequoia Applied Technologies came on as the engineering partner responsible for the full stack: frontend, backend, pipelines, and reference data infrastructure. This was not a consulting engagement that ended with a slide deck. Sequoia wrote the code, shipped the releases, and kept the system running as it grew.
The first chunk of work was building pipelines for the company's existing product lines. Some pipelines were net new. Others were replacements for older scripts that had become difficult to maintain. Sequoia also integrated the pipelines with an internal orchestration platform the client used across other product lines, so scientists and ops teams would see a consistent interface regardless of which assay type they were ordering.
The second chunk was the reference database layer. Assay design pipelines need to validate probe sequences against known variants and genomic annotations. The client relied on public data sources like NCBI dbSNP for genetic variants and curated databases for genomic annotations. Sequoia built ETL tooling to ingest these sources, some arriving as JSON dumps, others as flat TSV files, and normalize them into a queryable store. The reference layer had to stay current as upstream sources published updates, and it had to be fast enough that pipelines were not bottlenecked waiting on lookups.
The third chunk was the web application itself. Scientists log in, see the product catalog, submit design requests with their target lists, and track orders through validation, pipeline execution, and file delivery. Operations teams get dashboards showing queue depth, job status, and error rates. The backend exposes RESTful APIs that serve the frontend, internal applications, and the pipeline execution layer.
Angular Frontend, .NET Services, and Pipeline Orchestration
The architecture splits into three tiers: a single-page web application, a services layer that handles business logic and job management, and a pipeline execution layer that runs the actual bioinformatics work.
The web interface is built in Angular. Scientists see a product catalog, submit custom design requests by uploading target lists, monitor order status, and download generated design files. The UI is persnickety about input validation because catching errors early saves pipeline cycles and support tickets later.
The backend runs on .NET with a layered services architecture. One RESTful API serves the web UI and other internal applications. A second API handles pipeline execution, picking jobs off the queue, dispatching them to workers, and updating status as stages complete. SQL Server stores order metadata, manifest files, and audit logs.
Sequoia built ETL tooling to ingest reference data from NCBI and other public sources. The variant reference data arrives as JSON. Genomic annotations come as TSV. The ETL munges these into a normalized schema and loads them into a store optimized for the access patterns pipelines actually use.
Deployments follow a staged path from Dev to Test to Production with CI/CD automation through Jenkins. Automated tests run on every build. As the reference database grew, Sequoia introduced tiered storage to push older snapshots to cheaper archival tiers, keeping costs under control without sacrificing the ability to reproduce historical designs.
Over time, Sequoia handled framework upgrades to keep the stack current, moving from older versions of Angular and .NET to their modern successors. These migrations happened without breaking production or forcing scientists to relearn the interface.
Self-Service Assay Design with Full Audit Trail
The platform replaced a manual, email-driven process with a self-service system that scientists actually use. Turnaround from request to delivery dropped because jobs run automatically instead of waiting for someone to kick them off by hand. Support load dropped because scientists can see their order status without asking.
The audit trail gives the client something they did not have before: the ability to trace any design file back to the exact inputs, reference database versions, and pipeline runs that produced it. That traceability matters for customers in regulated industries who need to document how their assays were created.
For Sequoia, this engagement demonstrated the kind of end-to-end platform work the firm does for life sciences companies: not just writing code, but owning the whole stack from frontend to pipelines to data infrastructure, and keeping it running as the business grows.
Questions About Bioinformatics Platform Development
What is a custom assay design platform?
An assay design platform lets researchers specify the molecular targets they want to study, then generates the probe sequences and experimental layouts needed to run those experiments. The platform validates inputs against reference databases, checks for specificity and sensitivity issues, runs design pipelines, and delivers files that can be sent to manufacturing. Without such a platform, design work requires manual coordination between scientists, bioinformaticians, and production teams, which slows turnaround and introduces errors.
What types of research does a genomics assay platform support?
Genomics assay platforms support a range of molecular analysis applications including cancer research, pharmacogenomics, agrigenomics, and infectious disease studies. The platform handles design requests for different assay types, validates them against the appropriate reference databases, and generates outputs tailored to each application. A platform that handles multiple assay types allows a life sciences organization to serve a broader range of customers from a single system.
Why do assay design platforms need reference databases?
Reference databases provide the ground truth that assay designs are validated against. This includes databases like NCBI dbSNP that catalog known genetic variants and curated databases of genomic annotations. Building and maintaining these reference layers is a significant engineering task because the source data arrives in different formats, updates on different schedules, and must be normalized into a single queryable store that pipelines can hit without manual intervention.
What did Sequoia build for this life sciences client?
Sequoia Applied Technologies served as the end to end engineering partner for the platform. The work included building pipelines for multiple product lines, developing ETL tools to ingest reference data from NCBI and other public sources, creating a web application for scientists to submit design requests and track orders, and building the backend services that run pipelines, manage job queues, and deliver design files. Sequoia also handled ongoing framework upgrades and introduced tiered storage to control costs as the reference database grew.
What technology stack did Sequoia use for the assay design platform?
The frontend is a single page application built in Angular. The backend runs on the .NET framework with a layered services architecture exposing RESTful APIs. One API serves the web interface and other internal applications. A separate API handles pipeline execution and status updates. The database layer uses SQL Server for metadata and manifests. Deployments follow a staged Dev to Test to Production path with CI/CD automation through Jenkins and automated test coverage.
What kind of life sciences software does Sequoia Applied Technologies build?
Sequoia Applied Technologies is a Santa Clara software engineering firm that works with life sciences, healthcare, and biotech companies on platforms that handle sensitive data, complex workflows, and regulatory requirements. Engagements include bioinformatics pipeline development, laboratory information systems, clinical data platforms, and custom software for genomics and diagnostics companies. The firm has delivered similar work for NGS visualization platforms, clinical trial systems, and medical device software.