DEV Community

Sherdil IT Academy
Sherdil IT Academy

Posted on Originally published at academy.sherdil.org

Stop Comparing Cloud Institutes by Vibes: A 20-Line Scoring Script

If you are choosing a cloud training institute in Pakistan, you have probably done the same thing most people do: visited a few websites, sat through a couple of sales calls, and picked the one that felt most convincing. The problem is that "felt convincing" is exactly what a sales call is engineered to produce. So here is a way to take your own judgment out of the sales room and put it on paper: a weighted scorecard, in about twenty lines of Python.

The idea

Pick the factors that actually predict a good outcome, decide how much each one matters to you, rate every institute on each factor from 0 to 5, and let arithmetic do the comparison. The weights are the important part, because they force you to admit what you actually care about before you see any brochure.

WEIGHTS = {
    "hands_on_labs":     30,  # share of class time spent in real environments
    "instructor_current": 20,  # still working in the field, not just teaching
    "cert_alignment":    20,  # syllabus maps to the exam's published objectives
    "mentor_ratio":      15,  # realistic student-to-mentor support
    "placement_support": 10,  # interview prep, referrals, portfolio review
    "cost_transparency":  5,  # exam fees, retakes, extras stated up front
}

def score(ratings: dict) -> float:
    """ratings: each factor rated 0-5. Returns a score out of 100."""
    return sum(WEIGHTS[k] * ratings[k] / 5 for k in WEIGHTS)

institute_a = {"hands_on_labs": 4, "instructor_current": 3, "cert_alignment": 4,
               "mentor_ratio": 2, "placement_support": 3, "cost_transparency": 5}
institute_b = {"hands_on_labs": 2, "instructor_current": 5, "cert_alignment": 3,
               "mentor_ratio": 4, "placement_support": 4, "cost_transparency": 3}

print(score(institute_a))  # 69.0
print(score(institute_b))  # 67.0
Enter fullscreen mode Exit fullscreen mode

The two institutes above are hypothetical, chosen to show a close call. On these weights, A edges out B by two points, mostly because it is lab-heavy.

Why the weights matter more than the ratings

Now change one thing: suppose you know you struggle without hands-on guidance, so you shift weight from labs to mentoring.

WEIGHTS.update({"hands_on_labs": 20, "mentor_ratio": 25})

print(score(institute_a))  # 65.0
print(score(institute_b))  # 71.0
Enter fullscreen mode Exit fullscreen mode

The ranking flipped, and nothing about either institute changed. Only your priorities did. That is the whole point of the exercise: the "best" institute is not a fact about the institute, it is a fact about the match between the institute and you. Any published ranking, including ours, is answering someone else's weights.

How to fill in the ratings honestly

A score is only as good as the evidence behind it, so rate from things you can verify, not from things you were told.

  • Hands-on labs: ask for the syllabus and count lab sessions against lecture sessions.
  • Instructor current: check their recent work history, not their five-year-old bio.
  • Cert alignment: the syllabus should reference the exam's real domains by name.
  • Mentor ratio: ask for the actual batch size and how many people answer questions, not the marketing number.
  • Placement support: ask for a specific recent example, a role and a timeline, not a percentage.
  • Cost transparency: get exam fees and retake costs in writing before you pay.

If an institute will not answer one of these, rate that factor a 0 or 1. Evasion is data.

One limit of any scorecard

A spreadsheet cannot see delivery quality. Two institutes with identical syllabi can teach very differently, so treat the score as a way to shortlist two or three candidates, then sit in on a trial class before you decide. If you are shortlisting specifically in Karachi, what to look for in a top IT training academy there, including the red flags worth walking away from pairs well with this script, since it covers the local details a generic rubric will miss.


*I ran this kind of comparison across the cloud training market in Pakistan and wrote up the full findings, institute by institute, over here: Best Cloud Computing Training Institutes in Pakistan 2025-2026: Comparison.

What weight would you put on hands-on labs versus mentoring? Drop your numbers in the comments and see whose ranking changes. 👇

Top comments (0)