Look at the work behind the profile.
GullyMentor learners finish each level with reviewed practice projects in one of nine tracks: AI, machine learning, data analytics, data science, cloud, DevOps, cybersecurity, AI automation and business analytics. If you are hiring at entry level, tell us the role and we will suggest learners whose reviewed work matches.
What you get.
Learners who have completed a reviewed project: the problem, their contribution, the quality checks and the changes made after feedback. You see the work, not a claim about it.
There is no fee to employers, no placement contract and no obligation on either side. GullyMentor does not publish private candidate profiles and does not claim employer partnerships.
How it works.
Send the role, the required skills and the working arrangement through the form below, or by email. The team replies, usually within one working day, with whether learners with matching reviewed work are available. Learner details are shared only with the learner's agreement.
Interviews, selection and terms are entirely yours. Programme participation and a portfolio are inputs to your evaluation, not a guarantee of suitability.
What the work looks like.
The project briefs on this site show the standard a learner's work is reviewed against: a defined problem, synthetic or licensed data, quality checks and an honest account of limitations.
An example practice brief in each track.
Mentors set briefs on synthetic or appropriately licensed data. These are illustrative, and each learner's own project is set by their mentor to fit their level.
| Track | Example practice brief | Illustrative output |
|---|---|---|
| AI and Generative AI | Campus information assistant | A conversation prototype, source-linked answer sheet and failure log. |
| Machine Learning | Demand prediction baseline | A notebook with baseline and model predictions plus an error chart. |
| Data Analytics | Sales performance dashboard | A dashboard, calculation dictionary and reconciliation sheet. |
| Data Science | Customer behaviour investigation | An EDA notebook and a qualified findings memo. |
| Cloud Computing | Small website hosting plan | An architecture diagram, runbook and cost checklist. |
| DevOps | Tested application pipeline | A repository, pipeline file and recorded pass/fail evidence. |
| Cybersecurity | Local web security review | A sanitised report with scope, evidence, impact and retest steps. |
| AI Automation | Enquiry triage workflow | A process map, workflow demonstration and exception queue. |
| Business Analytics | Retail margin review | A KPI workbook and a five-slide management review. |
How the work is reviewed in each track.
- AI and Generative AI
- Bring the source material, prompt versions and failed examples. Feedback can focus on unsupported claims, ambiguous instructions and whether your evaluation actually measures the intended task.
- Machine Learning
- Check the split before the score. Feedback should examine data leakage, baseline choice, metric selection, error patterns and whether another person can reproduce the result.
- Data Analytics
- Review definitions and source totals before chart styling. Check duplicate rows, join behaviour, date filters, denominator choices and whether the dashboard answers a specific question.
- Data Science
- Inspect the question and sampling process, then the code. Examine missingness, confounding variables, uncertainty and whether the written conclusion goes beyond the available evidence.
- Cloud Computing
- Trace a request from DNS to the application and its storage. Check access permissions, exposure, backup assumptions and whether the teardown removes billable resources.
- DevOps
- Demonstrate a failed check as well as a successful release. Review reproducibility, permissions, secrets, deployment approval and whether rollback restores the intended version.
- Cybersecurity
- Start with permission and scope. Review reproducibility, whether the evidence supports the severity, what information must be redacted and how the remediation can be tested.
- AI Automation
- Walk through both a successful run and a failed run. Review the trigger, field mapping, duplicate handling, approvals and the information recorded in logs.
- Business Analytics
- Check whether the measure has a clear owner, definition and decision. Review the recommendation's assumptions, the alternatives considered and the information still needed.
Send your requirements.
The team replies by email, usually within one working day. No fee, no contract, no obligation. Prefer email? hello@gullymentor.com.
Tell us what good work looks like in your team.
Send the role and the skills; the team replies, usually within one working day.
Send your requirements