Outcomes

Horizon AI Coach

One coach per student, available at 2 a.m. before a drive. Mock interviews, aptitude sprints, GD labs and resume rewrites — all scored.

1,946 sessions this week

System · Placement EngineNorth star · Placed % of eligible batch24% of modelled movement

One trainer cannot run 3,660 mock interviews. The coach runs unlimited voice mocks, scores against a rubric recruiters actually use, and escalates only the students a human should see.

Open the Placement Engine →

Median mock score

68 / 100

Target 80 / 100 before drive season

Accountable

Training & Placement

Dean — Training & Placement

Students onboarded

4,218

Sessions this week

1,946

Avg score gain

+17

Over 6 weeks

Interview-ready

1,124

Weekly streak

68%

Students active 4+ days

Coaching volume

Sessions completed per week

▲ 62%

W1W2W3W4W5W6W7

Batch skill profile

Average score across all active students

DSA72

System design54

Aptitude81

Communication63

Domain depth69

Behavioural77

Readiness ladder

Every student sits on exactly one rung with a defined exit criterion

Level 1 — Foundations

1,120

Exit: Aptitude 60%+, resume v1

Level 2 — Core prep

1,486

Exit: 80 DSA problems, 3 mocks

Level 3 — Interview ready

1,124

Exit: Mock score 75+, 2 projects live

Level 4 — Tier-1 ready

488

Exit: System design 70+, referral routed

Drill library performance

Runs, average score and six-week gain

Drill	Runs	Avg	Gain
DSA rapid round	4,820	71	+14
System design basics	1,640	58	+19
HR & behavioural	3,110	76	+9
Aptitude sprint	5,240	79	+16
Group discussion lab	1,288	64	+12
Resume rewrite	2,960	68	+21
English fluency	2,410	61	+18

Cohorts

Final year — CSE486 students

Weakest: System design · avg improvement +18

Final year — ECE302 students

Weakest: Aptitude speed · avg improvement +11

Pre-final — AI&ML214 students

Weakest: Communication · avg improvement +24

MBA — Batch 2027168 students

Weakest: Guesstimates · avg improvement +15

Sample coaching transcript

System design mock, final-year CSE

Horizon Coach

Walk me through how you'd design a seat-booking system for campus events.

Student

I'd start with a table of events and seats, and lock a seat when someone books.

Horizon Coach

Good instinct. What happens when two students hit book in the same millisecond?

Student

Maybe a transaction... I'm not sure how to prevent the double booking.

Horizon Coach

That's the gap. Read up on optimistic locking and row-level locks — I've queued a 12-minute drill and one practice problem for tonight.

Recent sessions

Rithika M. · Mock interview — SDE

Strong DSA, hedges on trade-off questions.

Farhan A. · Resume rewrite

Projects lack measurable outcomes; rewrite queued.

Nandini P. · Aptitude sprint

Ready for tier-1 shortlists; move to system design.

Joel Thomas · Group discussion

Interrupts peers; assigned two GD labs this week.

Interventions the coach triggered

Automatic escalations to trainers and mentors

Streak break

312 students

No session in 7 days before a drive — WhatsApp nudge with a 10-minute drill.

Score plateau

148 students

Mock scores flat across three attempts — escalate to a human trainer.

Overconfidence

64 students

High self-rating, low mock score — assign a recorded panel mock.

Breakout

89 students

+25 points in four weeks — fast-track to tier-1 track and mentor pairing.

Signals at a glance

The four numbers this module is reviewed on weekly

Weekly active

3,120

Avg session

26 min

Questions served

1.4M

Human calibration

Monthly

Coached students vs interview clearance

AI coach usage against technical round clearance

Q1Q2Q3Q4Q5Q6

Weekly active students · 3120Clearance % · 59

Practice minutes by skill

184k minutes logged this term

184k

minutes

DSA6234%
Aptitude4122%
System design2212%
Communication3418%
HR / behavioural2514%

Score lift by intervention

Average points added to interview readiness

+6.4

Daily drills

+5.1

Mock feedback loop

+3.7

Weak-topic replays

+2.2

Peer battles

-2.8

Streak breaks

Net +14.6

Weekly sessions by mode

Voice coaching now the majority mode

Voice mockText drillCode round

Practice volume vs score gain

Bubble size is cohort headcount

Weekly practice minutes →↑ Readiness gain

Daily users (88/84)3x weekly (66/68)Weekly (42/46)Sporadic (22/24)Dormant (8/9)

Skill gap heat by branch

Higher means larger gap to recruiter bar

DSAAptitudeDesignCommsDomainCSE3428524022ISE4234604430ECE5840724836Mech7852845630MBA4630582634

Most improved students

Readiness delta over eight weeks

Student	Branch	Start	Now	Sessions	Delta
Rohit K	CSE	44	82	96	+38
Ananya D	ISE	51	86	88	+35
Faizan M	ECE	38	71	102	+33
Sneha P	AI/ML	58	89	74	+31
Kiran B	Mech	33	62	81	+29

Coaching integrity

Feedback is only useful if it is trusted

Feedback accepted by student82%

Rated useful within 24 hours

Score agreement with human panel88%

Calibrated monthly against faculty mocks

Streak retention (4 weeks)57%

Biggest single lever on outcomes

Escalation to mentor91%

Stuck students routed within one day

Risk register

What could take the number backwards

Gaming the drills

35% likely · Inflated readiness

Mitigation: Randomised proctored checkpoints

Streak collapse in exam weeks

70% likely · -14 pts avg

Mitigation: Auto-lighten cadence during internals

Over-reliance on AI feedback

40% likely · Weak human interviews

Mitigation: One human panel mock per month is mandatory

Forecast scenarios

Modelled outcomes with confidence, reviewed each month

Base80% conf.

59% clearance

3,120 weekly active students

Push59% conf.

66% clearance

Daily streaks at 70%

Stretch41% conf.

72% clearance

Full batch onboarded with proctored mocks

Outcome ladder

Each rung has a current number, a target, an owner and a date

Stage	Today	Target	Owner	By
Weekly active students	1,946	3,000	Coach agent	Oct
Students at Level 3+	1,612	2,400	Training cell	Dec
System design coverage	54	72	CSE/ISE HoDs	Jan
Human escalations closed	61%	95% within 5 days	Trainers	Ongoing

Levers

Where the movement actually comes from

Session frequency2.4 → 4

Students at 4+ sessions a week gain 16 points a month.

System design54 → 72

The single blocker between Level 3 and tier-1 shortlists.

Communication63 → 78

Fluency drills move this faster than any other intervention.

Streak retention57 → 80

312 students break streak in the fortnight before a drive.

Operating cadence

What the agent runs, and when

Nightly

Drill assignment

Each student gets tomorrow's 12-minute drill based on today's weakest rubric line.

Weekly

Plateau scan

Three flat mock scores triggers a human trainer escalation.

Pre-drive

Company-shaped mock

Rubric switched to the actual recruiter's interview pattern.

Data feeds

Nothing here is hand-entered

Rubrics

Recruiter interview patterns from 41 drives

Student level

PlacementIQ readiness ladder

Mentor notes

MentorSphere session homework

Unit economics

What a point of movement costs

Annual spend

₹26L / year

Unit cost

₹92 per completed mock

Payback

One semester

Contribution

24% of goal

+21 points average gain on resume and fluency drills