At Light My Career, transparency is everything. We believe career decisions should be backed by facts, not guesswork. Our platform aggregates data from the most reputable government and institutional sources and uses grounded Artificial Intelligence to interpret it for you.
1. Career & Salary Data
We use a Verified Hybrid Data Model to ensure you get the most accurate financial outlooks possible.
The Verified Database (Federal Data)
For established career paths, we pull directly from the U.S. Bureau of Labor Statistics (BLS). This is the gold standard for economic data in the United States.
Median & Hourly Wages: We use the Occupational Employment and Wage Statistics (OEWS) survey to determine median salaries and location based pay.
https://www.bls.gov/oes/
Growth Projections: We use the Employment Projections (EP) program to show you the 10 year outlook for specific jobs.
https://www.bls.gov/emp/
The AI Advantage (Niche & Emerging Careers)
For modern, niche, or emerging career paths where specific BLS data does not yet exist, we use OpenAI models (gpt-5.6-terra, gpt-5.5 and gpt-5.4-mini) to bridge the gap.
Estimation: We use advanced AI models to estimate salary ranges based on real time market signals.
Manual Verification: We do not rely on AI blindly. If a career is in our Verified List, it has been manually reviewed by our team to ensure the data is accurate.
Career wage and projection data covers the United States labour market. If you are studying in Canada, use the Canadian earnings figures on our school and program pages for local context.
2. U.S. Institutional Outcomes & Student Data
We look beyond the brochure to show you the real world results of a degree.
U.S. College Scorecard
We utilize the complete dataset from the U.S. Department of Education's College Scorecard. This provides us with comprehensive, standardized data to help you measure potential return on investment:
Post-Graduate Earnings: We show median earnings of graduates broken down by specific major, so you can see financial outcomes for your intended field.
Admissions Data: We display acceptance rates, SAT and ACT score ranges, and enrollment statistics to help you gauge your chances.
Cost: We show published tuition alongside average net price by family income band, so you can see what students actually pay rather than the sticker price.
Student Body & Retention: We provide transparent data on demographics, retention rates, and typical debt loads.
https://collegescorecard.ed.gov/data/
3. Canadian College & Program Data
We cover 235 Canadian public and private institutions and roughly 14,000 programs across all ten provinces and three territories.
Canada has no national equivalent of the College Scorecard. No federal agency publishes earnings, admission rates or net cost for individual Canadian institutions. Rather than leave the country out or fill the gap with guesses, we use the two national datasets that do exist and we label clearly what each one measures.
Program Earnings (Statistics Canada, 2021 Census)
For each program we show the median employment income of people aged 25 to 64 who worked full year full time, whose highest credential is in that field of study at that credential level, in that province. This comes from Census table 98-10-0409-01, using the 2019 income year to avoid pandemic distortion. It is what allows us to separate Civil Engineering from Mechanical Engineering rather than showing one figure for all of engineering.
https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=9810040901
School Earnings (Statistics Canada, ELMLP)
For each school we show an estimate built from Statistics Canada table 37-10-0280-01, which links postsecondary student records to tax records and reports median employment income five years after graduation by province, field of study and credential level. Those provincial medians are averaged across the programs the school offers. It is a modelled estimate, not a measured result for that school's graduates.
https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=3710028001
What we do not show for Canadian schools
Acceptance rates, SAT and ACT ranges, admission chance estimates and net cost after aid are not published for Canadian institutions in any comparable national dataset. We leave those fields empty rather than estimate them.
All Canadian figures are in Canadian dollars and adjusted for inflation.
4. Official Program & College Summaries
Every program and college page includes a written overview built only from the institution's own website.
How it works
We use Perplexity's sonar-pro model with search restricted to the school's official domain. For a program summary the model reads up to four official pages, typically the program page, the curriculum or course list, the admissions requirements and the tuition schedule. For a college summary it reads up to eight.
https://docs.perplexity.ai/
What we ask it to cover
Program pages: program overview, length and format, curriculum, hands on learning such as co-op, practicum or capstone work, admissions fit, career outcomes, important details, and the list of official sources.
College pages: test policy, application deadlines, tuition, and general institutional details.
The rules we enforce
Official school pages only. We do not pull from rankings sites, forums, aggregators or marketing partners.
No invented details. If the model cannot confirm something on an official page, it is left out.
Closest real match. If a program is not offered under the exact name you searched, we summarise the school's nearest equivalent program and say so, rather than telling you the program does not exist.
Every summary lists the official pages it was built from, so you can verify anything yourself.
Freshness
Program summaries refresh roughly every 90 days and college summaries roughly every 180 days. Institutions change tuition, deadlines and course structures between refreshes, so always confirm details with the school before making a decision.
5. Match Scores & Admission Estimation
This is where we turn raw numbers into personal insights. We use OpenAI models (gpt-5.6-terra, gpt-5.5 and gpt-5.4-mini) to analyze the data specifically for your profile.
Match Scoring: We use AI to generate a precise Match Score that indicates how well a school or career aligns with your specific goals. For colleges, this calculation incorporates median earnings, career trajectory estimates, and profile fit. For careers, we evaluate how well the role lines up with your professional objectives and preferences.
Admission Estimation: By analyzing your academic stats against the historical admission data from the College Scorecard, our AI provides an estimation of your admission chances, helping you build a balanced college list. Admission estimates are available for U.S. schools only, since Canadian institutions do not publish comparable admissions data.
6. The AI Career Coach
Our AI Career Coach is not a standard chatbot. It is a data backed advisor powered by OpenAI models (gpt-5.6-terra, gpt-5.5 and gpt-5.4-mini).
Where you start the conversation shapes what it knows. Open the Coach from a career and it receives that career's verified BLS wage and projection data. Open it from a college or a specific program and it receives that school or program's data, including earnings, cost and the official summary described above. Your profile comes with you either way.
Live Web Search: The Coach can search the web during a conversation, so it can look up current details such as application deadlines, program changes or recent labour market news rather than relying only on what the model was trained on.
Grounded Answers: Because the Coach is handed our verified figures at the start of the conversation, the salary, earnings and cost numbers it quotes come from the sources on this page rather than from the model's memory.
Personalized Guidance: Whether you are comparing two schools or analyzing a career path, the Coach uses your specific profile and our verified data to give you a tailored answer.
One thing to know: the Coach does not have our entire catalogue in front of it. If you ask about a school, program or career you have not opened, it answers using web search and general knowledge rather than our verified database. For the numbers we stand behind, open the page first.