eCAP — Campus Acceleration Track
A 4-week, campus-ready path from complete beginner to interview-ready: 20 weekday training days, Saturday doubt-clearing clinics, four cumulative assignments and a capstone — covering Salesforce development with Claude and Gemini used deliberately, not as a crutch.
Program shape
20 weekday training days (Monday–Friday, 4 weeks). Every day runs 2 hours theory → 15-minute break → 45 minutes hands-on practice. Every Saturday adds a 1-hour doubt-clearing clinic.
2h45m / weekday
2 hours theory, a fixed 15-minute break, then 45 minutes of guided hands-on practice — same shape every day.
4 Saturday clinics
One 1-hour doubt-clearing session per week. No new theory — just unblocking, right when that week's assignment is released.
4 weekly assignments
Done outside class, supported by the Saturday clinic — cumulative, building toward a capstone rather than four unrelated exercises.
Capstone by Day 19
The four assignments build on each other, so by Day 19 a student has an almost-complete small app to talk about in interviews.
Day 19 → Day 20
19 days of theory and hands-on practice, then Day 20 is a dedicated Interview Preparation day — no theory/practice split, the whole session is interview-focused.
Beginner → interview-ready
Scoped for a campus cohort starting from complete beginner, covering the Salesforce + AI/Agentforce topics that actually show up in interviews.
The four assignments
Each assignment builds on the last, so by Assignment 4 every student has one small, working, end-to-end app to talk about in an interview.
| Assignment | Released | Due | What it covers |
|---|---|---|---|
| Assignment 1 | Sat., end of Wk 1 | Mon., start of Wk 2 | 2–3 related custom objects, a 2-profile security scenario (OWD + role hierarchy + sharing rule), one Flow automation, one report + dashboard. |
| Assignment 2 | Sat., end of Wk 2 | Mon., start of Wk 3 | A bulk-safe Apex trigger + test class (incl. a 200-record bulk test) and one async job (Batch or Queueable), built on Assignment 1's data model. |
| Assignment 3 | Sat., end of Wk 3 | Mon., start of Wk 4 | An LWC + Apex controller (with error handling) surfacing Assignment 1/2 data, plus a CSV data-migration exercise upserted with an External ID, deployed via the CLI and committed to GitHub. |
| Assignment 4 (Capstone) | Day 17 | Day 19 | One Agentforce agent (subagent + custom action) grounded on the running app's data, tested with a 5-utterance suite, a clean Code Analyzer scan, and a README explaining the architecture. |
Self-grading: after each assignment, paste your work into Claude with "Here's my [assignment]: [paste]. Grade it and point out anything a Salesforce reviewer would flag." This isn't a substitute for instructor feedback in the Saturday clinic — it's a first pass so clinic time goes to real blockers, not basic mistakes.
Design choices behind the 4-week format
What changed to fit a 30-day / 3h-a-day curriculum into 20 days / 2h45m-a-day — and what didn't.
Why 20 days, not 30
A campus cohort runs on a fixed calendar. This track keeps every core Salesforce + AI/Agentforce topic that shows up in interviews, but compresses the 30-day version by merging closely related topics into single days and moving repetition-heavy practice out of class.
Where the missing practice time went
In-class practice is 45 minutes, shorter than a full hour. The difference is absorbed by four weekly assignments — supported by the Saturday clinic — rather than by cutting concepts.
No separate project phase
There isn't a multi-week runway for a solo Mini Project or a 5-role Team Project in a 1-month program. Instead, the four assignments are cumulative, standing in for a project when it's time to talk about it in an interview.
Trimmed to a brief mention
Headless 360 (newest, least foundational) and Copado/enterprise DevOps tooling (useful to know exists, not to go deep on) are folded into Day 18 as short overviews rather than full days.
Merged, not cut
Data Model I/II, Apex Classes + Testing, Agentforce's two deep-dive days, and DevOps Center/Code Analyzer + Best Practices each collapse from two days into one — the content is the same, the pacing is tighter.
Stayed at full depth
Security Model, Apex triggers & bulkification, LWC + Apex integration, Integration & Data Migration, Agentforce, and AI-assisted-development critical review — the topics interviewers and real projects actually probe.
How to use this plan
Every Monday–Friday session follows the same shape.
| Segment | Duration | What happens |
|---|---|---|
| Theory | 2 hours | Concepts, live demos, AI-tool walkthroughs |
| Break | 15 minutes | Fixed midpoint break — every day, same slot |
| Hands-on practice | 45 minutes | Guided exercise applying that day's concept |
Doubt-clearing, not new theory
Four 1-hour sessions, one per week. Pure doubt-clearing — no new theory. Each Saturday is also when that week's assignment is released, so students can bring questions about it immediately.
Saturday 4 is the program's last session: open Q&A on the full 4-week program, with students bringing blockers from Assignment 4 or anything across all 4 weeks. Time permitting, it also runs an informal extra mock-interview slot for anyone who wants another round.
What to have ready
- A free Salesforce Developer Edition org (developer.salesforce.com/signup)
- A Trailhead account, linked to the dev org
- Salesforce CLI + VS Code with the Salesforce Extension Pack
- AI tool accounts: Claude (claude.ai) and Gemini (gemini.google.com) — this plan scopes to these two only
- A GitHub account for version control
Interview Preparation
A dedicated day, not split into theory/practice — the whole session is interview-focused.
From assignments to resume bullets
Turn Assignments 1–4 into 5–8 resume-style bullet points, e.g. "Migrated data using Data Loader upsert with External ID matching and de-duplication"; "Built and tested an Agentforce agent with a custom Flow-based action, validated via a Testing Center suite." Ask Claude to phrase these the way a Salesforce job posting or interviewer would expect.
Technical question bank (sample)
- Explain the difference between a Lookup and a Master-Detail relationship, with a scenario for each.
- What does "bulkification" mean, and what anti-pattern breaks it at 200 records?
- Walk through the Data Cloud pipeline from raw ingestion to an activated Segment — where does Identity Resolution fit?
- An AI coding tool generates a trigger calling a method that doesn't exist in your org. Describe your process for catching this before deployment.
- What's the difference between a subagent and an action in Agentforce? What role does the Einstein Trust Layer play?
- Why is upsert with an External ID generally safer than a plain insert for a migration you might re-run?
Mock interview rounds
- Pair up and run two 15-minute mock technical rounds using the question bank above, then switch.
- Ask Claude to interview you Socratically about your own capstone: "Ask me 5 probing questions about my design decisions, one at a time."
Behavioral / HR round
- Practice "tell me about a project you've worked on" using the STAR method (Situation, Task, Action, Result), grounded in your capstone.
- Prepare a short, honest answer to "how did you use AI, and how do you know the output was correct?" — this is a real question interviewers ask now.
What's next
This program builds toward Salesforce Certified Administrator, Platform Developer I, and Agentforce Specialist — worth pursuing once you're through this track.
How exercises are labelled
Under 15 minutes, a single concept.
15–25 minutes, combines 2–3 concepts.
The full 45-minute slot, a near real-world scenario.
Use deliberately, not as a crutch
| Tool | Best for |
|---|---|
| Claude | Explaining concepts deeply, architecture/design decisions, code review, documentation, debugging logic errors, red-teaming AI-generated code |
| Gemini | Quick boilerplate/scaffolding, research, comparing approaches, multimodal help (diagrams/screenshots), brainstorming test cases |
Golden rule: try to solve it yourself first (or at least attempt an outline), then consult AI to check, refine, or unblock. Never paste AI output into your org without understanding every line — this is a learning plan, not a copy-paste plan.
Guardrails for using AI throughout
- Attempt every task yourself first, even just an outline — AI should accelerate learning, not replace it.
- Always review AI-generated Apex for: bulkification, CRUD/FLS/sharing enforcement, and governor limits.
- Never deploy AI-generated code — or an AI-generated agent action — you can't explain line-by-line.
- Use AI to generate alternatives and critiques, not just first drafts — asking "what's wrong with this?" is often more valuable than "write this for me."
- When using Claude or Gemini for Agentforce work, remember the actual guardrails your agent relies on in production are the Einstein Trust Layer and your own scoping — don't treat AI-approved instructions as a substitute for testing in Testing Center.
- Before any data migration, back up the target data first, test the load in a sandbox, and prefer upsert with an External ID over a blind insert.
- Static analysis (PMD/ESLint/Flowtest via Code Analyzer) is a gate, not a suggestion — an unresolved high/medium finding blocks calling the work "done," same as a failing test.
Apply through the Programs menu
eCAP cohorts are listed alongside enpropeL's other programs and events. Open the Programs / Events menu, find the current eCAP cohort, and apply from there.
- Login required to apply
- Free for students at partner campuses
Submit a professional services enquiry
Bringing eCAP to your campus starts as a Professional Services enquiry with eCAP pre-selected as the service. No account needed to submit it — a form opens either way, logged in or not.
- No login required to submit
- enpropeL's team follows up within 48–72 hours
Bring eCAP to your campus.
A structured partnership that gives your students free access to internships, jobs, mentors and the Learning Hub, with your institution's branding on the journey.