AI Integration Developer

AI Integration Developer

Location: Hybrid at Lemont, IL (50% of the time onsite) Duration: 6 Months with possibilities of extension

Only W2 candidates are eligible for this position. Third-party or C2C candidates will not be considered.

Description:

Client, through the NST division, is seeking an AI Integration Developer contractor experienced with the development, enhancement, and maintenance of AI-assisted workflows supporting scientific data.

Some common tasks that will be performed include working with scientific domain experts to understand the location and structure of their data, designing methods to extract/develop metadata, implementing a strong dataset taxonomy to ensure consistency, orchestrating the migration of datasets for computational analysis using approved LLMs, etc.

Scope:

Client is seeking an engagement with an experienced candidate with qualifications in the listed technologies to aid the NST division and CNM User Facility in work to automate/standardize the process of collecting and storing data for AI training and collaborative ventures.

Objectives:

The overall goal of this assignment is to deliver a functional and impactful workflow to aid research scientists across the NST division with data management, including, but not limited to:

• Identifying location of data and the methodology to determine which data is to be managed through the workflow
• Filtering, sanitizing, normalizing and/or otherwise preparing the data for storage/sharing
• Packaging the data with AI-enabling metadata into an HDF5 format that can be used for long-term storage
• Creating a mechanism to migrate and inventory data for easy retrieval and/or sharing
• Building an interface/workflow that is accessible and easy to use by the research staff
• Develop documentation of solutions
• Construct all components, artifacts, etc. using acceptable tools and techniques, good coding practices and accessible documentation

Tasks and Delivery:

• Three projects have been identified that require data manipulation, sanitization, orchestration and automation.
• During the first 2-3 weeks, engagements with the scientific domain experts will provide details on understanding their datasets, locations and desired outcomes.
• Once the details have been identified by the domain scientists, work will begin through an iterative development process to create a compatible workflow, using existing or new tools to achieve the desired outcomes. All tools, software and related materials will require compliance with local cybersecurity and IT standards as well as documentation of operation. Development under revision control is required.
• Delivery to the domain scientists of an MVP within the first 90-days of the project is expected.
• The MVP need not address all of the key deliverable outcomes but must demonstrate the core of the workflow desired.
• Further refinement, refactoring and feature addition is expected beyond the 90-day window.
• Once the domain scientists have accepted the solution, the candidate will be assigned a new set of domain experts and a new workflow to create.
• It is expected that each contractor will have no fewer than two active development tasks.

Required Background, Knowledge, Skills and Abilities:

Knowledge/Skill/Ability:

• A strong working understanding of AI and LLMs
• Ability to develop efficient prompts and structured workflows with AI integration
• Familiarity with AI workflow orchestrators (e.x. Dify)
• Familiarity with relational databases (e.x. MySql/MariaDB, Oracle, etc.)
• Working knowledge of relevant programming/scripting languages (e.x. python) for automating tasks
• Working knowledge of revision control systems (e.x. Git)
• Working knowledge of CI/CD pipelines as implemented in services like gitlab/github for testing and deploying solution changes automatically
• Familiarity with RESTful web services
• Strong skills in Linux (Debian and RHEL) CLI administration, development and general use
• Familiarity with GUI development either through web or python applications
• Strong written and oral communication skills providing the ability to develop requirements for tasks through interactions with technical experts in the field

Background:

• Associate or bachelor’s degree in computer science, information technology, system administration, or a closely related field, or equivalent experience required
• 2 or more years of software development and/or system administration

Work Location:

• The primary location of contract performance will be primarily accomplished at the contractor’s facility. Between 20-40% of the individual’s time will be required on-site.

Work Schedule:

• Mostly off-site schedule (up to 80%) with 40 hours per week. Schedule to be negotiated with hiring manager.

Technical Direction of Work:

• PSE’s IT Manager will manage the workload, but the tasks will be driven by the domain experts in NST/CNM.

Task and Time Reporting:

• Time will be reported weekly, and performance will be measured against assigned tasks, scheduled maintenance and productivity.

Government-Owned Property:

• The work will be completed using Government Furnished Equipment (GFE) complete with all required cyber protections, software licenses and tools needed to complete the work under the contract.
• The device will be made available to the contractor at the outset of the contract, once all cyber-related training is complete. At the completion of the contract term, all provided GFE shall be returned promptly to the Lab.

Security Requirements:

• As part of the contract, access will be granted to datasets and information systems secured with restrictive permissions.
• Access to such systems is explicitly for the contractor only.
• Accounts used for access may not be shared with others.
• Datasets and other information assets may not be shared, copied or transferred out of Laboratory-owned systems, including GFE noted above.
• No clearance is required.

Computer Protection Program:

• The contractor(s) may have root access to government computers and/or Business Information Systems.
• The contractor(s) shall adhere to all policies and procedures of the client Computer Protection Program, must not bypass any procedures established to protect data, applications, hardware, or communications at client, must maintain a work environment that will satisfy audit, privacy, and protection requirements, and must report any findings of inadequacies to the technical contact and the client Cyber Security Representative.