Report Examples¶
This is what ControlAI actually produces. The data is invented but realistic; the layouts are simplified so it's clear what goes where. Below each example is a breakdown of what every part means. All the examples below are shown in Russian; in the system, reports are generated in the group's language — Uzbek, Russian, or mixed — chosen when the group is created.
1. Lesson Card¶
This is what the quality-control office (ichki ta'lim sifati nazorati bo'limi), a dean, or a head of department sees when they open a specific class session; an instructor sees the same kind of cards for their own sessions. Below are two cards for the course "Higher Mathematics (Oliy matematika)": a lecture and a practical class. A standard session is 80 minutes.
Lecture¶
┌───────────────────────────────────────────────────────────────────────────┐
│ Higher Math · Lecture: Definite Integral ● Ready │
│ Instructor: Karimov A. (assoc. prof.) · 2nd-year stream · Room 214 │
│ Thu 09:00–10:20 · actual: 09:02–10:18 │
├───────────────────────────────────────────────────────────────────────────┤
│ Instructor talk 82% Student talk 6% Silence 12% │
│ Questions to stream 9 Started +2 min Ended −2 min │
├───────────────────────────────────────────────────────────────────────────┤
│ Lesson DNA: │
│ ██████████▓▓██████░░████ (lecturer · questions · students · silence) │
├───────────────────────────────────────────────────────────────────────────┤
│ Highlights: │
│ ✓ 09:58 Question block to the stream — 7 questions in 6 minutes │
│ ✗ 09:24 26-minute monologue with not a single question │
├───────────────────────────────────────────────────────────────────────────┤
│ Report: The lecture on "Definite Integral" was delivered in full, │
│ on schedule. A high share of lecturer talk is normal for a │
│ lecture; recommendation — break up the long monologue with │
│ short questions more often. │
│ Observation-form score: 8 / 10 │
└───────────────────────────────────────────────────────────────────────────┘
Practical Class¶
┌───────────────────────────────────────────────────────────────────────────┐
│ Higher Math · Practical: Definite Integral ● Ready │
│ Instructor: Rasulova N. (instr.) · Group 210-21 (year 2) · Room 118 │
│ Thu 10:30–11:50 · actual: 10:33–11:50 │
├───────────────────────────────────────────────────────────────────────────┤
│ Instructor talk 44% Student talk 39% Silence 17% │
│ Questions 27 Started +3 min Ended on time │
├───────────────────────────────────────────────────────────────────────────┤
│ Lesson DNA: │
│ ██▓▓░░████▒▒██▒▒░░██▒▒ (instructor · questions · students · silence) │
├───────────────────────────────────────────────────────────────────────────┤
│ Highlights: │
│ ✓ 11:05 Board work — students explained solutions for 12 minutes │
│ ✓ 10:52 "Q&A" streak: 11 consecutive student responses │
│ ✗ 11:38 Independent work cut short by the bell — no wrap-up │
├───────────────────────────────────────────────────────────────────────────┤
│ Report: The definite-integral practical had active group │
│ participation: 27 questions, four board-work episodes, a healthy │
│ talk-time balance. Silence at 17% is mostly independent problem- │
│ solving, which is normal. │
│ Observation-form score: 8 / 10 │
└───────────────────────────────────────────────────────────────────────────┘
What's what:
- Header — the course and session type, instructor, stream/group, room, the scheduled time and the actual time (from the recording): the session really ran bell-to-bell, not just "on paper."
- Metrics — the session's key numbers (what they mean — see the Metrics chapter). Norms depend on the session type: 82% lecturer talk in a lecture is normal; the same 82% in a practical class is worth a conversation. Adjusting for type is currently up to the reader; automatic calibration of benchmarks is on the roadmap.
- Lesson DNA — a colored ribbon showing how time was split (instructor / questions / students / silence).
- Highlights — good (✓) and problem (✗) moments tied to a timestamp.
- Report — a short text written from the ready-made numbers.
- Observation-form score — how closely the session matched the university's own observation form.
And if a session didn't happen at all, the card looks like this (short):
┌───────────────────────────────────────────────────────────────────────────┐
│ Higher Math · Practical · Group 210-22 · Room 119 │
│ Toshmatov B. (senior instr.) · Thu 06/25, 08:30–09:50 ⚠ LESSON NOT HELD │
│ Recording was retrieved for the session window; no speech was detected. │
│ Flag auto-raised in the dean's and head of department's summaries. │
│ Not included in the billed volume. │
└───────────────────────────────────────────────────────────────────────────┘
This answers a question that used to require someone checking hallways in person: did each session actually happen — now it's confirmed objectively, from the recording.
2. Daily Summary¶
This is what a dean and the quality-control office see on the daily dashboard — every session in the faculty at a glance, with problems surfaced at the top.
Summary for Thursday, June 25 · Faculty · 46 sessions · 28 instructors
Student talk 34% ↑ Questions 19 Punctuality 88% ↓ Held 43 / 46
⚠ NEEDS ATTENTION
● Room 119, 08:30 — session not held (group 210-22, Toshmatov B. (senior instr.))
● Room 122, 13:30 — session not held (group 310-21, Nosirov U. (instr.))
● Room 305 — camera offline: the 11:00 session recording wasn't retrieved, needs an audit
● Toshmatov B. (senior instr.) — started the 10:30 session (group 210-24) 18 minutes late
◐ Karimov A. (assoc. prof.) — a 26-minute monologue in the lecture with not a single question
★ Rasulova N. (instr.) — best session of the day in the faculty: 39% student talk, 27 questions
SESSIONS OF THE DAY
08:30 Higher Math · 210-22 · Toshmatov B. (senior instr.) ················ not held
09:00 Higher Math (lecture) · Karimov A. (assoc. prof.) ·················· ready 82% lecturer talk
10:30 Higher Math (practical) · 210-21 · Rasulova N. (instr.) ············ ready 39% student talk
10:30 Higher Math (practical) · 210-24 · Toshmatov B. (senior instr.) ···· ready started +18 min
10:30 Transportation Organization (seminar) · Yusupova G. (assoc. prof.) · processing
11:00 Electrical Engineering (lecture) · Ahmedov Sh. (assoc. prof.) ······ no recording (room 305)
13:30 Materials Science · 310-21 · Nosirov U. (instr.) ··················· not held
...
What's what:
- Header — the faculty's daily summary numbers with trend arrows, including "Held X / Y" — how many sessions actually took place. Here, 43 of 46 scheduled sessions were held: two are flagged "session not held" (the first is the same card from Section 1), and one has no recording at all — the camera in room 305 is offline. These are different things, and the system doesn't conflate them: "not held" means there's a recording but no class happened (excluded from the billed volume); "no recording" means an equipment problem — the session's status isn't confirmed, and the ticket goes to the technical team.
- "Needs attention" — the system automatically surfaces what matters: sessions not held, late starts, equipment problems, problem moments within sessions — and standout results. Color/icon signals the severity. Note: Toshmatov B. (senior instr.) had two sessions that day — the first (08:30, group 210-22) wasn't held, the second (10:30, group 210-24) happened but started 18 minutes late.
- Sessions of the day — a list with a status and one key number each.
3. Summary for the Rectorate¶
One level up is the university-wide aggregate: not individual sessions, but a comparison of faculties and departments. Usually it's the vice-rector for academic affairs who works with it and brings the finished summary to the rectorate meeting — the rector doesn't need to sit in the panel to see the picture. With hundreds of sessions a day, this is the only realistic format: not looking at everything, but at where the deviation is. The summary can be built for any period — day, week, semester; a week is shown here.
University · week of June 22–28 · 1,640 sessions · 12 faculties
Held 96% ↑ Student talk 33% → Punctuality 91% ↑ Not held 62 ↓
FACULTIES THIS WEEK SESS. NOT HELD STUDENT TALK PUNCT. TREND
Transportation Management 214 3 37% 95% ↑
Rolling Stock 198 5 34% 92% →
Transport Infrastructure 186 9 31% 88% ↓
...9 more faculties
DEPARTMENTS OF THE WEEK
★ "Transportation Organization" — 38% student talk, not a single session
missed
★ "Electrical Engineering and Electronics" — average rating has been rising
for the fourth week running
▼ "Materials Science" — 6 sessions missed this week, 74% punctuality
▼ "Higher Mathematics" — the monologue share in practicals is rising: 61% →
70%
(despite some strong individual sessions)
What's what:
- Header — the whole university in one line: thousands of weekly sessions distilled into four numbers with trend arrows against last week.
- Faculties — comparable figures for each faculty: where sessions missed exceed the norm, where students stay silent, where punctuality is dropping. It's the same data as the dean's daily summary (example above), rolled up to a week and the whole university.
- Departments of the week — the best (★) and the attention-needing (▼) departments across faculties: a reason to recognize some and question others, backed by numbers.
- A single document — the summary exports as one piece: ready-made material for the rectorate meeting.
4. Personal Report for the Instructor¶
This is what an instructor sees in their own account — only about themselves, framed as help, not an inspection.
┌─────────────────────────────────────────────────┐
│ Rasulova N. (instr.) — my profile │
├─────────────────────────────────────────────────┤
│ My rating 8.6 ↑0.2 │
│ Student talk 39% Questions 27 │
│ Punctuality 94% │
├─────────────────────────────────────────────────┤
│ My goal: bring student talk up to 45% │
│ Currently 39% ▓▓▓▓▓▓▓▓░░ — solid progress │
├─────────────────────────────────────────────────┤
│ What to improve: leave 5 minutes at the end of │
│ class to review independently solved problems │
└─────────────────────────────────────────────────┘
What's what:
- Own metrics and rating — the instructor's personal statistics; in this personal format — with goals and tips — only they see it, while leadership sees the numbers in their own summaries.
- Goal — self-set or agreed with the head of department, with automatic progress tracking.
- What to improve — a tip for growth, not a "grade from above." The rating is internal analytics for development, not a substitute for official certification.
5. Observation-Form Compliance (Completed Form)¶
This is what the quality-control office or a head of department sees and checks — essentially an open-lesson (ochiq dars) observation form filled in for every session, not once or twice a year by a committee.
Session: Higher Math · Practical · Rasulova N. (instr.) · Group 210-21
Observation-form score: 8 / 10 (confidence: high)
CRITERION VERDICT CONFID. EVIDENCE
Started and ended on time ◐ Partially high start 10:33 (+3 min)
Topic and goals announced ✓ Yes high "Today we're reinforcing the
Newton—Leibniz formula through
problems"
Link to lecture material ✓ Yes medium references to the stream's morning
lecture
Student engagement ✓ Yes high 27 questions, 4 board-work
episodes
Final wrap-up, conclusions ✗ No medium wrap-up cut short by the bell
Classroom discipline ? Needs a person low (video-based criterion)
TOPIC COVERAGE (Definite Integral)
✓ Newton—Leibniz formula ✓ change of variable
◐ integration by parts ✗ improper integrals
What's what:
- 1–10 score — not a "bare number," but paired with a confidence level.
- Criteria — from the university's own observation form; for each one: a verdict, how confident the AI is, and a quote as evidence from the session. The AI honestly flags disputed items as "needs a person."
- Topic coverage — what's been covered from the syllabus (✓), partially covered (◐), not yet covered (✗).
- The quality-control office confirms or corrects — and the system learns from the corrections.
The key takeaway from this chapter: ControlAI's output isn't a "wall of numbers" — it's clear cards: a breakdown of a lecture and a practical class, a flag for a session not held, a daily faculty summary for the dean, a university-wide summary for the rectorate, an instructor's personal dashboard, and a filled-out observation form. Every part is labeled and explainable.
→ Next: 10. Value for Everyone — what all of this gives each role.