01 · Principles
Fit, not a universal ranking
Drempo does not rate which employers are “good.” We estimate characteristics that may matter differently to different engineers. High work intensity may be energizing to one person and a serious compromise to another. A large, structured organization may feel supportive or constraining. The quiz makes those preferences explicit.
Employer-wide patterns are only a starting point. A team, manager, location, contract, product phase, and specific role can differ substantially from the broader organization.
02 · Fit dimensions
Twelve tradeoffs with plain-language meaning
Compensation
How strongly the employer appears to reward technical talent relative to its market.
Work-life balance
The likelihood of predictable personal time, based on public role and work-pattern evidence.
Flexibility
Remote, hybrid, schedule, and location flexibility where the work allows it.
Stability
Organizational durability, business breadth, and predictability of employment conditions.
Career growth
Visible development programs, internal mobility, and breadth of technical career paths.
Autonomy
How much room technical employees may have to shape work and decisions.
Work intensity
A spectrum from measured and predictable to fast, demanding, and deadline-driven.
Technical depth
Access to difficult engineering problems, specialized systems, research, or advanced tools.
Company structure
A spectrum from lean and fluid to large, process-rich, and highly structured.
Mission
How central a tangible public, environmental, health, or infrastructure purpose is to the work.
Collaboration style
A spectrum from independent deep work to highly cross-functional, team-intensive work.
Risk and upside
A spectrum from steady compensation to entrepreneurial risk, equity, and rapid-growth upside.
03 · Matching algorithm
Deterministic, weighted, and intentionally modest
Quiz answers create two things: an importance weight from 0–5 and, where relevant, a preferred position on a 1–5 spectrum. Compensation, balance, flexibility, stability, growth, autonomy, technical depth, mission, and upside are benefit signals: the model rewards stronger evidence in proportion to how important the user says that signal is.
Intensity, company structure, and collaboration style are directional. Their fit is calculated as 1 − |employer value − preferred value| ÷ 4. A person who wants a measured pace is not pushed toward the most intense employer, and a person who wants a lean team is not automatically sent to the largest organization.
The weighted dimension fit is combined with modest industry and discipline relevance. We then rescale the result into a deliberately bounded percentage. The number is a communication aid, not scientific precision, probability of happiness, or a promise about employment.
Ties are resolved consistently by employer name. There is no randomness, popularity boost, sponsorship adjustment, or manual override in personalized placement.
04 · Research and sources
Public evidence first
We prefer official careers and benefits pages, company reports, public filings, government information, reputable news, and legally usable public job information. Each employer profile links to its current official career and company sources and shows the last research date.
We do not scrape Glassdoor, copy Indeed reviews, reproduce copyrighted employee reviews, use login-gated sources, invent benefits, infer protected characteristics, or use protected classes in matching. Employer claims remain attributed to the employer.
Browse the source registry05 · Editorial standard
Publisher-level accountability
Research is published by Drempo Research, the editorial unit of the Drempo organization. We use an organizational byline rather than invent named writers, credentials, or a team that does not exist.
Deep content separates documented public facts, Drempo interpretation, general role-based observations, confidence, and limitations. Comparisons do not declare a universal winner. The specific role, manager, program, site, and current job posting remain the final unit of diligence.
06 · Confidence
Uncertainty is part of the result
Each subjective dimension has a confidence label. High means multiple direct, current public signals support the broad characteristic. Medium means the evidence is useful but incomplete or role-dependent. Limited means evidence is thin, indirect, or especially likely to vary by team.
Confidence describes the evidence, not the employer. A limited-confidence estimate is a prompt to investigate, not a negative score.
07 · Updates and corrections
Claims are dated and challengeable
Guides show published and reviewed dates. Priority employer profiles show reviewed dates and multiple sources. Substantive source or interpretation changes update the reviewed date; silent spelling or layout corrections do not imply that the research itself was rechecked.
Correction requests should identify the exact page, statement, current primary source, and source date. Drempo evaluates the evidence, corrects material errors, and keeps editorial conclusions independent of employer requests or sponsorship.
08 · Automation and AI assistance
Tools can assist; sources remain accountable
Software and AI-assisted tools may support source discovery, consistency checks, drafting, link validation, or structured-data generation. They are not treated as sources. Factual claims must trace to the listed public evidence, and editorial responsibility remains with Drempo Research.
The quiz itself is not a generative-AI assessment. It uses a deterministic rules-based model, and individual answers remain in the user’s browser.
09 · Limitations
What Drempo cannot know
Public information cannot reliably describe every manager, team, workload, compensation offer, project, or location. Employer policies change. Job postings can lag practice. Private companies disclose less. Scores therefore help create a shortlist and interview questions; they should never replace direct diligence.
Drempo does not provide legal, financial, recruiting, or employment advice. It does not predict an offer, performance, advancement, or happiness.
10 · Permanent independence rule
Sponsorship never affects a match
An employer cannot purchase a higher characteristic estimate, a place in personalized results, or removal of a legitimate tradeoff. Any future sponsored profile will be labeled and separated from the matching model. The model version used at launch is drempo-fit-v1.
Methodology last reviewed: August 25, 2026.