All too often, students cram career development into the last year, month, or minute of their academic experience. They crave highly personalized ideas about jobs that fit their major, dreams, and skills from a handful of career advisors who are deluged by hundreds or thousands of students. On the other side of the table are increasingly finicky employers using software tools to screen applicants. And everyone is wondering about the impact of AI.
Honestly, who knows what the impact of AI will be on jobs and careers? But we do know that AI can break the bottleneck in Career Services: it is available 24 hours a day, no appointment needed. It can handle thousands of students at a time. It is endlessly patient with students’ ruminations about interests, majors, and jobs. It can keep track of real-time hiring trends, wages, and skill requirements. It can make sound, data-informed suggestions about jobs that alumni really get and propose more creative ideas that may fit a student’s skills and interests. And it can draft resumes, prepare students for interviews, and help with all the other schmutz of modern job hunting. Importantly, AI can inform and free up human advisors to provide better data and more personal attention to students in a tough job market.
All of this matters – a lot. Federal and state regulators, accreditors, and boards of directors are zeroing in on employment outcomes, and wages may soon limit access to federal financial aid. Alumni who feel they received good career advice are much more likely to believe that their investment of time and money was worthwhile, making them more loyal, more likely to donate, and more likely to recommend the institution to their friends. Importantly, applicants and parents want to know how you will help students get jobs. They expect human advisors, as they should. In addition, colleges and universities can now offer state-of-the-art systems that are well-informed, scalable, cost-effective, and available 24 hours a day.
The CIP-SOC Bridge: Connecting Education to Employment
Sound career advice must correctly link academic programs and jobs. While this seems straightforward – accounting majors become accountants – in practice, most people do not end up working in the fields their major would suggest. History or Philosophy undergraduates seldom become historians or philosophers. They both often become lawyers; Philosophy majors often become software developers, but History majors don’t. Gray DI’s AI career advisor (CoCo Careers) uses more than 30 million career profiles to provide students and advisors with data-informed insights on what people actually do with their degrees.
Introducing the Gray DI CIP-SOC Crosswalk
Every academic program in the United States is classified under a six-digit CIP code (e.g., 14.0901 for Computer Engineering). Every occupation tracked by the Bureau of Labor Statistics has a SOC code (e.g., 15-1252 for Software Developers). A CIP-SOC crosswalk identifies the jobs that are associated with a given academic program.
Given a sound CIP-SOC crosswalk, a student who says “I’m majoring in Information Science” can immediately see which occupations their program prepares them for, pay in their region, and skills employers need. Unfortunately, most crosswalks are well-intended fictions, made up by economists in conference rooms. They usually depend on the NCES crosswalk, which, according to NCES, is “not based on actual empirical data”.¹
Using over 30 million alumni records on degrees earned and jobs held, Gray DI developed its own CIP-SOC crosswalk. It provides a much broader and more accurate perspective on the link between academic programs, occupations, wages, and skills. While traditional crosswalks usually suggest 5-10 occupations for a program’s graduates, Gray DI’s alumni data reveal that graduates actually pursue dozens or hundreds of occupations; often, the top occupations are overlooked by a traditional crosswalk. Using this data-informed crosswalk ensures that students get sound advice about career options. In particular, it shows that there are well-paid jobs for liberal arts graduates. For example, software developer is one of the top 15 occupations for Philosophy graduates and Anthropology majors – and it pays over $100,000 per year.
CoCo’s Implementation
CoCo operationalizes the CIP-SOC crosswalk through integration with a data warehouse that combines federal classification data with real-time labor market intelligence. When a student in CoCo Careers asks, “What jobs can I get with my major?”, the system doesn’t rely on the AI’s general knowledge — it queries Gray DI’s crosswalk to retrieve specific, verifiable occupation matches with current job postings, salary ranges, and skill requirements.
DATA GROUNDING VS. AI OPINION
This distinction is critical. A general-purpose AI chatbot might suggest careers based on its training data — which may be outdated, regionally inaccurate, or simply fabricated. CoCo’s career advice is grounded in Gray DI’s extensive data and analysis, ensuring that salary ranges, job growth projections, job postings, and occupation-to-program mappings are verifiable and up to date.
Persistent Career Memory
Career development is not a single conversation — it is a relationship that evolves over semesters and years. For example, a first-year student may discuss their interest in renewable energy with a Career Advisor, then take an environmental science course during sophomore year, and complete a sustainability internship in their junior year. Typical large language models – and humans – are likely to have forgotten this narrative by the time a student needs help during their senior year. Or, the language model may remember absolutely everything the student entered, including that unfortunate session after an all-nighter. CoCo Careers maintains persistent memory across sessions, building a cumulative understanding of each student’s interests, skills, experiences, and aspirations. Importantly, it also enables the student to edit the memories, so only the relevant information is retained. This deals not only with off-topic screeds, but it also allows the student to modify their profile as their academic and career interests shift. The edited memories enable the career AI to provide adaptive and increasingly personalized guidance over time:
Interest tracking: “Last semester, you mentioned an interest in data analytics. Have you considered the new Business Intelligence certificate offered this spring?”
Skill evolution: “You’ve been working with Python in your data science courses. Based on current job postings in your target area, adding SQL and Tableau would significantly strengthen your profile.”
Goal alignment: “You mentioned wanting to stay in the Tampa Bay area. Here are the leading local environmental consulting employers with their current hiring trends.”
This memory is capped and managed to prevent unbounded growth while preserving the most career-relevant information. The history is private to each student and inaccessible to professors, other students, or institutional administrators. Using the history, CoCo builds a career profile that students can edit and share as they see fit.
Multimodal Career Engagement
Students engage with career guidance differently depending on their context, comfort level, and learning style. CoCo Careers provides several modalities:
Text-Based Career Chat
Text is the primary interface and is currently the most popular medium for students. Using text messages, a conversational AI can answer career questions, provide labor market data, discuss career pathways, and help with professional development. The Career Coach AI operates with its own personality and prompt engineering, creating a companion-like experience rather than a clinical advising session.
Voice-Based Career Coaching
Some students find it easier to “think out loud” about career aspirations than to type them. For these students, CoCo offers immersive, personal, voice-based interactions. Voice enables deeper, more reflective conversations — the kind students have when they’re genuinely uncertain about their future. It enables CoCo to simulate employment interviews, allowing students to prepare for the wide range of personalities and questions they may encounter. It coaches them to keep answers focused, well-spoken, and aligned with employer interests.
Structured Career Assessments
Not all career guidance should be open-ended. CoCo includes structured questionnaires — skills assessments, interest inventories, and exercises to clarify priorities — that help students who don’t know where to start. The results of these assessments feed into the persistent memory, informing future career conversations.
AI-Assisted Resume Building
One of the most tangible outputs of career guidance is a resume. CoCo can generate professional resumes using the student’s profile, skills, experiences, and career targets — then render them as polished, employer-ready PDF documents. It can quickly tailor the resume to align with specific job postings – a critical capability in a world of AI resume readers that scan for specific keywords.
The Workforce Alignment Imperative
The integration of career intelligence into the academic experience is no longer a nice-to-have — it is increasingly a regulatory and accreditation requirement. Federal and state policymakers are demanding that institutions demonstrate tangible links between academic programs and employment outcomes.²
Gainful Employment Standards
The federal government’s “gainful employment” framework evaluates programs based on whether graduates’ earnings are sufficient to repay their educational debt. The CIP-SOC crosswalk allows institutions to predict the programs that are likely to run afoul of the new regulations.
The Apprenticeship Initiative
In 2026, the U.S. Department of Labor launched major initiatives to integrate AI skills into Registered Apprenticeships nationwide, focusing on embedding AI curricula into traditional trades, infrastructure roles, and emerging high-demand sectors.³ For universities, this creates both an opportunity and a challenge: how do degree programs align with these evolving apprenticeship pathways? The AI Career Advisor provides students with a productive way to experiment with AI and learn how to use it on the job.
Accreditation Pressure
Regional and programmatic accreditors are increasingly requiring institutions to demonstrate “career readiness” outcomes as part of their assessment. An AI platform that tracks the entire student journey from enrollment through career exploration provides the longitudinal data that accreditors demand — data that a disconnected career center visit in Year 4 cannot supply.
Current Limitations & Conclusion
Limitations
Employer feedback loops: The current system provides career intelligence to students but does not yet incorporate employer feedback on graduate preparedness. An ideal future state would include employer signals that inform both the career AI and the academic curriculum.
Alumni tracking: Post-graduation career outcomes are not yet tracked within the platform. Longitudinal data connecting academic engagement patterns to employment outcomes would dramatically strengthen the career intelligence model.
Conclusion
The gap between the classroom and the career office is an artifact of institutional organization, not a requirement of the student experience. When career intelligence is embedded in the same platform where students learn — grounded in structured labor market data, informed by persistent memory of the student’s academic journey, and delivered through modalities that match how students actually engage — the transition from education to employment becomes a natural extension of the learning process rather than an abrupt, underprepared pivot.
For a generation of students navigating an AI-transformed labor market, the institutions that bridge this gap will produce the graduates who thrive. The ones that don’t will continue producing the 70% who graduate without field-relevant employment.
References
¹ NCES / Learn-Work Ecosystem Library, “CIP-SOC Crosswalk,” 2024–2026.
² Georgetown University Center on Education and the Workforce, “Workforce Alignment in Higher Education,” 2025.
³ U.S. Department of Labor, “AI Skills Integration in Registered Apprenticeships,” dol.gov, 2026.
