AWS for Industries

Resilience builds lab supply chain tracking platform on AWS

We would like to acknowledge the valuable contributions of (Resilience) Bridget Smith, Senior Systems Engineer; Jonathan Rivernider, Lab Systems Engineer; Brian McNatt, Digital Research Sites Head; John Kerwin, Tech Head of Gene Therapy; Rachel Anderson, Specialist II, Supply Chain; (iOLAP) Alex Guerrero, Principal Solutions Architect; Marko Kos, Data Engineer.

Improving biomanufacturing operational efficiency is crucial for life sciences companies to expedite the delivery of essential medications to patients. Life sciences companies have more digital systems than ever before, but when these systems are disconnected from one another, it detracts from the efficient coordination of people, information, and work that is needed to bring new innovative medications to the market. The industry now looks to the cloud for tools and approaches to create serverless automation between in a manner that can scale.

Resilience is a biomanufacturing innovation company, offering a range of scalable, off-the-shelf biomanufacturing modalities for gene therapies, nucleic acid synthesis, protein purification, and more for leading pharmaceutical and biotechnology companies. In this way, Resilience empowers their clients to achieve their goals of bringing new therapeutics to market faster by effectively managing the analytical development, quality control testing, and good manufacturing practice (GMP) services required for scalable drug production.

Traditional contract development and manufacturing organization (CDMO) tech transfer processes present a number of challenges with consumable inventory management, data provenance, and consumable cost recovery when running experiments on behalf of clients. Typically, consumable costs are roughly factored into the overhead of the project, which can create significant issues in two critical areas. First, tracking exact consumables across the experiments they were used in is essential for data integrity and auditability of the production process. Without precise tracking, it becomes difficult to ensure that the right materials are used in the right experiments, potentially compromising the quality and reproducibility of results. Second, an imprecise charging methodology can lead to inefficiencies in chargeback processes for financial teams. Leftover reagents that were initially factored into overhead but not used effectively can result in wasted resources or, in some cases, complicated efforts to offer these excess consumables back to clients. For Resilience, these challenges underscored the need for a more accurate and efficient approach to consumable inventory management and cost recovery within CDMO operations.

To address these industry challenges, Resilience identified disconnected processes that had been carried out by three separate teams within the R&D department and built a comprehensive platform to remove manual steps. Specifically, these processes included the ordering workflow, the warehousing workflow, and the consumption workflow.

The platform leveraged an event-driven architecture with two native AWS services: AWS Lambda and Amazon EventBridge. With these services, they integrated two widely-adopted SaaS applications: the Benchling electronic lab notebook and inventory system and the Coupa purchasing system. This combination of technologies and services has allowed Resilience to increase operational efficiency and data integrity in their lab consumable usage. This provided traceability for over 1,000 unique orders in the first three months of use, which equated to over $1 million in previously untracked spend.

Automating Scientific and Financial Systems

Resilience used a combination of AWS serverless services, Lambda, and EventBridge to address the above challenges and achieve the desired business outcomes. They chose a serverless approach because it allowed their IT teams to focus on business outcomes more than managing servers. Serverless improves the cost effectiveness of the workload through consumption-based pricing, and improves performance through auto-scaling.

The end-to-end process is split into three workflows, managed by three distinct teams to provide logical separation of responsibilities (Figure 1 Top). Amazon EventBridge was used as the interface to automate transactions between systems, starting with orders that scientists place within Benchling (Figure 1 Bottom).

lab consumable workflows

Integration approach

Figure 1 Top. The lab supply chain platform is made up of three workflows, initiated through the Benchling GUI – Ordering, Warehousing, and Consumption.

Figure 1 Bottom. High-level architecture for the workflows.

Ordering Workflow

Scientific teams leverage Benchling workflows to initiate new consumables orders. These workflows feature an intuitive graphical user interface (GUI) where scientists input essential details, including their preferred vendor, item catalog number, desired quantity, and the associated project. This data is automatically transmitted to the Lab Support team, which is responsible for verifying approvals for requests and ensuring that corresponding Purchase Requests (PR) are recorded in Coupa, Resilience’s financial procurement system. During this stage in the workflow, the Lab Support team also has the option of updating item catalog information within Benchling, which serves as a central list of previously ordered items that is easily requested again.

The workflow is: new order placed by scientist in Benchling; notification sent to lab support team; event bus and eventbridge rule; update item catalog with new price and other relevant fields in Lambda; Benchling.

Figure 2. Workflow to request that new consumables orders be placed

Warehouse Workflow

Upon arrival of the item(s) onsite, the Lab Warehousing team then unboxes and barcode labels each item. The team enters the PO number from the packing slip into the Benchling workflow. All relevant ordering information from Coupa such as total price, billing units of measurement, and requester, are automatically associated with the Benchling item by leveraging the EventBridge integration with AWS. EventBridge triggers an AWS Lambda function, which makes a corresponding API call to Coupa. This information is then relayed back to Benchling through another simple API call, such that items are registered in the inventory management system, along with all relevant financial information. (Figure 3 Top)

After item creation in Benchling, the Lab Warehousing team validates that the physical items were received, prints barcode labels containing this information, places barcodes onto each item, and stocks in the storeroom where scientists retrieve it. (Figure 3 Bottom)

The Warehousing workflow at a high level is: Receipt Part number lookup, to Receipt PO Response, to Receipt Container Creation to outputs. The workflow for Receipt Part number lookup is: Receipt in Benchling, user enters POs, EventBridge Rule, write PO info to output Lambda, to Coupa and Benchling. The workflow for Receipt PO response is to validate manually that items are received. The workflow for Receipt Container Creation is: user initiates container creation in Benchling, EventBridge Rule, Lambda, Item registered in Benchling along with Coupa PO Information, team prints barcode and labels item.

Figure 3 Top. Overview of the process for receiving a package, registering it within the inventory system, and accounting for it in the financial system. Figure 3 Bottom. Overview of the process for labeling the material with a barcode and updating its storeroom location in Benchling.

Consumption Workflow

With items now barcode labeled and registered in Benchling’s inventory system, scientists can easily record their usage by simply scanning the items into their electronic lab notebooks, thus tracking their utilization across experiments. Leveraging a similar EventBridge trigger and Lambda function as in the previous workflow step, teams can accurately record consumable usage against specific client projects and automatically reduce the remaining item quantity. This meticulous tracking not only enables precise item usage logging for client chargeback purposes, but also maintains an up-to-date and accurate onsite inventory.

The consumption workflow is: Initiate workflow from within ELN entry in Benchling; EventBus and EventBridge rule; Consumption Lambda; Automatically decrement running inventory level per item in Benchling.Figure 4. Workflow for consuming materials within a client project.

Building Event-Driven Architectures for the Life Sciences

AWS Lambda and Amazon EventBridge are simple to get started with to create automation between systems.

Lambda is a serverless, event-driven compute service that lets you run code for virtually any type of application or backend service without provisioning or managing servers. As in the example here, Lambda can communicate with third party software-as-a-service (SaaS) applications like Benchling and Coupa to take actions automatically, such as decrementing inventory, updating PO information, or registering new inventory.

Lambda enables automation in a variety of use cases in life sciences organizations. You can launch Lambda from the appearance of new data files in Amazon Simple Storage Service (Amazon S3), to notify users that new data is ready for analysis. Lambda can let bioinformatics applications know to begin a data processing workflow. Lambda can be used to flag – and modify – data quality issues in real-time, as data is being collected.

To develop automation with Lambda, simply write script or code in your preferred language, such as python or Node.js, within the AWS Lambda console, test it, and click deploy. When deployed, the code becomes a scalable web application with memory, compute, and networking resources managed automatically.

Here is an example of python script in Lambda that is used to call Coupa to request all information relating to a certain PO, for an order which was received by the Warehousing Team. (Figure 5)

Example of Lambda code in the AWS Console Lambda interface. This function places a GET request to the Coupa API url. This will return information about a PO defined by the token.

Figure 5. Example of Lambda code in the AWS Console Lambda interface. This function places a GET request to the Coupa API url. This will return information about a PO defined by the token.

EventBridge is an event bus system on AWS which lets software systems create events, route events, and consume (take action on) events. Modern SaaS applications like Benchling use EventBridge to publish events, which can be used to launch workflows, such as Lambdas. This allows front-end applications like electronic lab notebooks (ELNs), laboratory information management systems (LIMS), manufacturing execution systems (MES), and customer relationship management (CRM) systems to become the “cockpit” that orchestrates downstream events automatically.

EventBridge rules match an incoming event and send them to a target for processing. In the example below, the EventBridge rule for PO creation launches the Lambda code shown above. In this manner of matching rules to Lambdas, life sciences organizations create automation to streamline repetitive processes and increase data integrity. (Figure 6)

The result of this matched event will be to call the target Lambda that executes the Coupa lookup function.

The result of this matched event will be to call the target Lambda that executes the Coupa lookup function.

Figure 6 Top. Example of an EventBridge rule in the AWS Console EventBride interface. This rule will be a match when a new event is created that meet the criteria of name = “Receipt (Coupa Lookup)”. Figure 6 Bottom. The result of this matched event will be to call the target Lambda that executes the Coupa lookup function.


With the Lab Consumables platform in place, Resilience is addressing key industry challenges of accurate electronic lab records and accurate client chargeback on consumable costs. The company is also continuing to onboard different sites and product lines to the platform and is seeking to further enhance user experience by including new features such as item cataloging and automation of low-cost transactions within Coupa. The platform showcases Resilience’s firm commitment towards modern cloud capabilities and lab automation.

Cloud events can help create automation in the life sciences, in an easy-to-write manner. Services like Lambda, EventBridge, and other AWS serverless technologies allow you to easily connect your mission critical applications like ELN, LIMS, MES and financial systems to scalable storage and computing on AWS. The AWS Life Sciences Solutions page details the best practices and case studies in labs, manufacturing, supply chain, and other parts of the value chain. For more information, contact your AWS representative.

Further Reading

Adam Mendez

Adam Mendez

Adam Mendez serves as the Head of Data Engineering and Architecture at Resilience, building upon his background in research and data science, particularly in next-generation sequencing and bioinformatics. Previously a lab scientist, he transitioned his career to creating cloud-based data platforms to accelerate research outcomes for pharmaceutical companies and manufacturing. Adam holds a Masters degree in chemical and biomolecular engineering from Johns Hopkins University.

Alissa Gordon

Alissa Gordon

Alissa Gordon is a Principal Account Manager at AWS where she works with Life Sciences Startups. She is passionate about helping customers build and scale on AWS. While at AWS she has also held positions as a Partner Success Manager and a Partner Development Manager. She has a BA from Wellesley College, and holds Masters degrees from both Harvard University and Georgetown University.

Lee Tessler

Lee Tessler

Lee Tessler, Ph.D. is a Principal Technology Strategist for the Healthcare & Life Sciences industry at AWS. His focus is on cloud architectures for modernizing R&D, clinical trials, manufacturing, and patient engagement. Prior to joining AWS, he launched products in the areas of bioinformatics, drug discovery, diagnostics, lab instruments, and pharma manufacturing. Lee holds a Ph.D. in computational biology from Washington University in St. Louis and Sc.B. from Brown University.