Revolutionizing Apprenticeship Programs with Turbine's Integration of LLM Framework

Tue. Aug 15, 2023 - by Josh Studl

This is the first of several posts to come that will highlight the work we are doing with LLMs and vector embeddings. We are excited to share our progress and hope you enjoy reading about it.

Revolutionizing Apprenticeship Programs with Turbine’s evolution of LLM Framework

Apprenticeships offer a unique opportunity for individuals to acquire valuable skills and education while earning a living. However, the complexity of the Registered Apprenticeship (RA) system poses challenges for employers and apprenticeship sponsors alike.

Turbine Workforce platform has taken a significant stride in addressing these challenges by wholesale integration of Large Language Models (LLMs), NLP transformers, vector databases into our tech stack.

Turbine platform is effectively evolving into a framework to connect customer knowledge base, and RA metadata with large language models, like GPT-4, GPT-3.5-turbo, Cohere, LLaMA. We are particularly interested in converting natural language into API calls or database queries. Skulls start to blow their tops around here when we talk about an LLM output that contain arguments to call functions. 🤯 🚀

The impact and upside is pretty huge for RA Sponsors and employers. Despite the clear benefits to employers and employees of RAs, the intricacies of the process often deter employers from engaging in the program, while apprenticeship sponsors struggle with labor and business regulatory compliance management.

With Turbine, participation in RA programs will be simplified and automated. Compliance with RA standards, federal requirements and alignment with regional workforce development systems will be the standard, not the exception.

Scaling Registered Apprenticeships can present challenges as well. These apprenticeships are meticulously designed to provide paid on-the-job training and comprehensive instruction, cultivating highly skilled professionals in specific occupations. These programs maintain rigorous standards that outline essential training and proficiency benchmarks apprentices must achieve to successfully complete the program. Typically, individuals accumulate a minimum of 2,000 hours of on-the-job learning, complemented by a recommended minimum of 144 hours of relevant instruction.

However, documenting OJT hours in a logbook can be burdensome, time-consuming, and often poorly executed. Consequently, evaluating skill progression and maintaining balance across the RA competency model often becomes subjective and lacks quantification.

One way we have introduced LLMs to customers is with VELA, our Voice-enabled logbook application.

VELA effectively addresses this challenge by providing apprentices with a tool to verbally record their just completed task into VELA. The LLM classifies, parses, and creates a text embedding of the transcribed record, which, upon approval by the OJT manager, populates the logbook. Employers gain unprecedented visibility into the work their apprentices are doing, and accountability is built into the managers’ approvals. The impact is a higher yield on the apprenticeship investment for employers.

RA Sponsors and employers can now utilize Turbine’s LLM to create training plans, manage apprenticeship agreements, and track compliance. Employers can also leverage Turbine’s LLM to participate in RA programs, gain insights on regulatory compliance, make data-driven decisions, and reach a broader pool of apprentices regardless of their location.

Turbine’s integration with LLM not only simplifies and cost-effectively ensures compliance but also provides a centralized source for all compliance management data. This reduces the time spent on administrative tasks associated with managing the RA program, allowing employers to focus on other responsibilities and work more efficiently and effectively. Additionally, there is improved coordination with the Regional Workforce Development System, aiding in self-evaluation of the program and expediting RA certification.

Turbine’s LLM framework is a game-changer for Registered Apprenticeships. This integration facilitates a smoother process for employers, apprenticeship sponsors, and beneficiaries, significantly reducing entry barriers for both parties. It paves the way for more Apprenticeships, benefiting employers by expanding the available labor pool and empowering apprentices as they gain valuable skills and knowledge while earning a salary.