Summary
• Tecsup’s students already had ChatGPT, but faculty were still building slides, quizzes, and activities across Forms, presentation tools, and Canvas.
• The institute was not going to replace that LMS. After a late-2024 pilot, it licensed Edcafe for 500+ faculty so AI would sit on course content and get shared from Canvas.
• A year later they renewed. 72% of 144 surveyed instructors said they save time each week.
Faculty at Tecsup were spending hours each week building slides, questions, and activities by hand. Many split their workflow across ChatGPT, Microsoft Forms, Canvas, and presentation tools.
Planning felt slow. Feedback in labs and theory courses ate time that could have gone to teaching. Students already had ChatGPT, so extra prep did not always show up in class.
The constraint was not that faculty needed more AI. Teachers needed activities and feedback in one place, tied to the course, without leaving Canvas.
If you lead a school or institute, this case study is how Tecsup instructors solved that for 500+ faculty, and what instructors reported after a year.
Background
| Location | Peru — campuses in Lima, Arequipa, and Trujillo |
| Institution type | Private, non-profit technical and professional education |
| Scope | 500+ faculty across mechanics, digital technology, mining, general studies, sciences, and more |
| Teaching modality | 81% in-person · 15% hybrid · 3% online (July 2026 survey, n=144) |
Tecsup is a private, non-profit institution founded in 1982 to train professionals for industry. Its programs apply technology in industrial operations and maintenance, with workshops and labs at campuses in Lima, Arequipa, and Trujillo.
It launched one of Peru’s first virtual campuses in 1999 and works closely with employers to keep programs aligned with industry needs.
Most Tecsup instructors teach in person, often as part-time faculty balancing prep with other work. Canvas as their LMS already anchored how courses were run; any new tool had to sit alongside that workflow, not replace it.
This large-scale rollout was led by Julio Monjarás, Academic Director of Tecsup Sur. He has more than 30 years of experience in educational leadership and teacher training in ICTs.
Monjarás said:
“Our priority at Tecsup has always been to close educational gaps and foster quality learning. We chose Edcafe and renewed our institutional license because we saw a tool that, instead of adding to the workload, gives time back to the teacher. It allows us to apply active methodologies and ensure that AI works on our own academic content, not on generic answers from the internet.”
The challenge: prep time, fragmented tools, and generic AI
At Tecsup, the constraint was not simply “teachers need AI.” Part-time faculty were prepping and supporting students across five separate tools. Students already had ChatGPT or Gemini, but those answers were often too broad for the course.
According to a July 2026 survey of 144 Tecsup instructors, the most common pains before Edcafe were:
- Creating materials took too long — 49%
- Too many separate tools (ChatGPT, Forms, LMS, etc.) — 33%
- Grading or giving feedback took too long — 22%
- Low student participation — 17%
- Hard to tell if students understood — 16%
A mechanics instructor described the logistics:
“I spent too much time coordinating materials between my LMS, presentation platforms, and interactive apps. With information scattered across tools, it was hard to create differentiated versions of the same topic for students who needed basic support and those who needed a bigger challenge.”
Juan G., a mathematics instructor, said creating materials to spark student interest “was quite exhausting” before Edcafe, while Jym, a part-time General Studies instructor, noted that giving feedback on entrance exams was difficult “because of time and the number of students.”
The institution needed a way to:
- Build activities and AI support in one place, tied to course content
- Run quizzes and feedback faster without rebuilding entire courses
- Give students a course-bound AI option when generic chatbots were unreliable
- Support part-time faculty who cannot spend a semester learning a new platform
How Tecsup addressed those challenges
After a late-2024 pilot, Tecsup bought 500 Premium licenses in July 2025 and trained faculty on the courses they were already teaching. Instructors did not rebuild their courses. Activities were shared as Canvas links so students stayed in the LMS they already used.
Long prep and too many tools → one place for materials and AI
Faculty who had split prep across ChatGPT, Forms, slides, and the LMS started building activities in Edcafe and linking them from Canvas.
88% of surveyed instructors had used Chatbot and 65% had used Quiz, the two tools that mapped to the most-cited pains: long prep and slow feedback.
Topic-scoped chatbot
Instructors upload weekly PDFs, lab sheets, or session content and configure a chatbot so students can ask doubts asynchronously or during class, without the teacher re-explaining the same point every time.
Gerardo L, who teaches electrotechnology and electrical metrology, uses the chatbot “to guide students, reinforce their learning, and help them understand the topics covered in class.”
YouTube Quiz and flashcards
A Mechanics instructor, assigns applied mechanics videos as pre-class material, then uses YouTube Quiz and flashcards so students review key formulas before lab:
“The biggest benefit is how it centralizes functions that normally require three or four different apps.”
Slow grading and feedback → faster checks and rubric-based grading
Quizzes and formative checks
Entrance exams, end-of-class comprehension checks, and Canvas-linked assessments replaced manual question-building for many part-time faculty. Students got feedback before the next class or lab, not after a weekend of hand-grading.
Dario F., a Tecsup instructor, uses Edcafe before class for “summative evaluations of the complementary information students must review before the session.”
Assignment Grader for written and lab work
Another group of instructors uploads rubrics for faster feedback on essays and lab submissions.
Dino T., who teaches drilling and blasting, uses Assignment Grader for written work: “The platform helps me generate rubrics, grade automatically, and give personalized comments to each student in a fraction of the time it took before.”
Low participation and hard-to-read understanding → in-class and pre-class checks
Quizzes, flashcards, and chatbots gave shy or overloaded students a way to engage without speaking up in a large technical class.
A mechanics instructor reported that after adding short quizzes and flashcards, students “feel more involved and stay on track during long theoretical explanations.”
Generic AI off-topic → course-bound chatbots and program routing
Instead of sending students back to open-ended ChatGPT, teachers limited AI answers to the sources they uploaded, or built program-level bots for continuing education.
Rene G., Coordinator of CPE in the Mechanics Department, went further and built program-level chatbots for PFR and CPE so participants could get answers routed to the right place without waiting for a coordinator.
Related reading: Dudley College’s six-month Edcafe trial - 80% of tutors said lesson planning got faster.
What changed: survey results from 144 instructors
About a year after the school-wide license began, Tecsup surveyed 144 of its 500+ licensed faculty.
72% said Edcafe saves them time each week.
82% reported at least one classroom gain from a multi-select list that matched the pains they’d named before:
- Faster feedback to more students
- More student participation
- Less manual prep
- Fewer apps per workflow
- Knowing which students had gaps before the next topic
In practice, feedback reached students before the next lab or exam, not after a weekend of grading. Students who’d normally stay quiet in a large technical class still completed quizzes and used chatbots to work through doubts. A short quiz at the end of a session showed instructors where to focus before moving on.
Marco M, a Tecsup instructor, put it simply:
“Edcafe is a very powerful and easy-to-use tool.”
What teachers hear from students
Student response was largely positive.
Dino T.’s students “greatly appreciate receiving grades and comments almost immediately, which lets them correct mistakes before the next evaluation.”
Gerardo L. also noted students were more willing to ask questions and more confident working at their own pace.
In IT-heavy programs, some students still preferred the commercial AI tools they already use. It’s worth factoring in, especially for programs where students are already heavy AI users.
Related reading: How Almuslim University used Edcafe to reduce math anxiety
Conclusion
A license renewal is a harder signal than a pilot. Pilots are optimistic. Renewals are evidence. Tecsup renewed after 500+ faculty, a full year, and a survey showing 72% save time each week.
They got there because the tool returned time without changing how faculty work. They kept Canvas as the student layer, trained on courses instructors were already teaching, and let part-time faculty adopt at their own pace.
For other technical institutes, the standard is simple: don’t ask faculty to change how they work. Pilot with real courses, license when it holds, train against what teachers already teach. That’s what separates a tool faculty actually keep using from one that fades after the pilot.