The future of college counseling

The best college advisor in the world is an AI.

College Signal is the first advisor that sees every data point — your student’s transcript, scores, activities, narrative — against the published admit-rate data and per-major modeling that drive every probability. No human advisor can hold this much context. No human advisor has read this many applications.

COLLEGE SIGNAL · REPORT · ARJUN PATEL · 2026
§01 · 1 of 11
Acceptance Probabilities
Per-school probability read
MIT (EECS)
Reach
9%
Stanford CS
Reach
6%
Carnegie Mellon SCS
Reach
14%
Georgia Tech CS
Reach
28%
U. Maryland CS
Likely
75%
+ 4 more schoolsGenerated in minutes
What's inside

Seven things no counselor can compute.

A counselor gives you experience, intuition, and pattern-match. The report gives you arithmetic — every probability conditioned on every lever the student still has time to pull. Here are excerpts from a real January 2026 report.

§ 1–10 of 11 (post-exec-summary)
SUBJECT: ARJUN P., LEXINGTON MA
STRATEGIC REPORT · GROUNDED IN SOURCE DOCUMENTS
§ 1
Acceptance probabilities

Every school, every probability, every confidence interval.

For each school, the report gives a point estimate, a 90% credible interval, and the four signals most responsible for pushing the number up or down. No more guessing which school is the real reach.

§ 1
School
Tier
Admit %
Arjun's prob.
90% CI
Top drivers
MIT
Reach
4.0%
9%
119
+ USACO Platinum
+ Harvard research co-author
+ First-gen STEM
− CS pool base 3%
Stanford
Reach
3.7%
6%
116
+ SAT 1540 above p75
+ Research depth
− Stanford CS among most competitive
Caltech
Reach
3.1%
6%
116
+ Math/physics AP scores
+ Technical fit very strong
− No Olympiad medal
Carnegie Mellon SCS
Reach
11.0%
14%
424
+ USACO Platinum load-bearing
+ Research topic aligns
− SCS pool ~5% base
UC Berkeley EECS
Reach
11.0%
14%
424
+ Academic profile above median
+ Research mentorship
− OOS penalty + EECS pool ~5%
Georgia Tech CS
Reach
17.0%
28%
1838
+ Stats above admit median
+ Strong fit for College of Computing
− OOS applicant pool competitive
UIUC CS
Reach
42.0%
23%
1333
+ Grainger fit strong
+ Research + USACO read well
− Direct-admit CS pool ~7%
U. Michigan CSE
Reach
18.0%
28%
1838
+ Above 75th% SAT
+ Rigor matches
− OOS cap on CS admits
Purdue CS
Likely
50.0%
80%
7090
+ Well above admit median
+ Demonstrated CS interest
≈ Fit signal sufficient
§ 3
Activities, ranked by weight

Not a list — a ranking, with the model's read of each signal.

Every activity is scored on the same internal scale admissions readers apply — tier, legibility, scale of impact, how well the signal fits the rest of the profile. The top two carry this application; the rest are supporting cast and need to be framed as such.

§ 3
#
Activity
Category
Tier
Weight
Model notes
01
USACO Platinum division qualifier
Academic Competition
T1
94
Top 4% nationally — rare and highly legible signal for CS programs.
02
Research: deep learning for protein structure prediction
Research
T1
91
Harvard Medical School mentor, 18 months, co-author on submitted paper. Tangible output sets it apart.
03
Founded + led ‘Code-for-Lexington’ non-profit (2 yrs)
Leadership · Impact
T2
78
12 contributors, 80 students across 4 elementary schools. Scale + outcomes need sharper framing in the activities list.
04
Varsity cross-country captain (junior, senior)
Athletics · Leadership
T2
71
Captaincy helps; no state-level placements blunts the tier.
05
Piano — ABRSM Grade 8 distinction, ~9 yrs
Arts
T3
58
Strong technical achievement. Reads as well-rounded rather than distinctive.
+ 7 more activities · Tiering reflects how selective admissions officers read the signal
§ 4.2 of 12
§ 4
Test score sensitivity

What a retake is actually worth — per school, per point.

Instead of one number (“your score is fine”), the report runs the admissions model against every realistic score scenario and reports the probability lift at each school. The answer is almost always counterintuitive.

§ 4
School
Current
SAT 1540
+20
SAT 1560
+40
SAT 1580
+60 (pract. ceiling)
SAT 1600
MIT
9%
10%+1
11%+2
12%+3
Stanford
6%
6%
7%+1
8%+2
Caltech
6%
8%+2
10%+4
12%+6
CMU SCS
14%
15%+1
17%+3
19%+5
UC Berkeley EECS
14%
16%+2
18%+4
20%+6
Georgia Tech CS
28%
30%+2
32%+4
33%+5
UIUC CS
23%
24%+1
26%+3
27%+4
U. Michigan CSE
28%
30%+2
31%+3
32%+4
Purdue CS
80%
81%+1
82%+2
82%+2
Model recommendation: Retake in November is not the highest-leverage move. A +40 composite lifts Caltech ~4pp and Berkeley ~4pp; the same 60 hours spent closing the senior-year coursework gap (§ 7) and attempting one more USACO contest projects a larger combined lift across the reach tier.
§ 5
ED / EA / REA strategy

Binding commitment is the single largest lever. Most students spend it wrong.

The report evaluates every allowable binding + non-binding early option against your portfolio as a whole — not just the school in isolation. Early-action at a school you'd get into anyway isn't a strategy; it's a wasted slot.

§ 5
Option
Constraint / note
RD prob.
ED/EA prob.
Portfolio impact
ED → MIT
MIT is REA, not ED. Not an option.
N/A
REA → MIT
Binding-lite single-choice. Blocks all other private EAs.
11%
16%
+2.8 pp overall portfolio
EA → Caltech
Non-restrictive. Stacks with other EAs.
14%
17%
+1.4 pp overall portfolio
EA → Georgia Tech
Non-restrictive. EA deadline Oct 15 — tight.
47%
54%
+1.1 pp overall portfolio
EA → UMich CoE
Non-restrictive; rolling admits favor early.
58%
66%
+1.4 pp overall portfolio
ED → CMU SCSRECOMMENDED
Binding. Available for SCS. Strongest single lever.
19%
31%
+4.9 pp overall portfolio

ED at CMU SCS is the single highest-leverage binding commitment for Arjun's profile. The model projects a 12 pp lift at CMU alone, which compounds into nearly 5 pp on portfolio expected-value because CMU sits at a steep section of the acceptance curve for his stats.

§ 6
Senior-year coursework

Three complete schedules, scored on probability lift and narrative coherence.

The transcript readers see is nearly set — but second-semester senior year and summer programming still move. The report proposes complete plans, not piecemeal advice, and tells you what each one signals to a reader.

§ 6
Spring ’26 baseline:
AP Calculus BC · completed 11th|
AP Physics C · completed 11th|
AP Computer Science A · completed 11th
MODEL PICK
OPTION A
PLAN A — Multivariable + Linear Algebra (dual enroll at Harvard Ext.) + AP English Lit + AP Stat + Data Structures (Lexington HS)
Strongest for CS-reach narrative
+3.1 pp reach avg
+Signals post-AP math competence directly
+Dual-enroll transcripts read as college-level rigor
+Data Structures closes one real gap
High workload Q1 — risk of B in AP Lit
Commute/schedule to Harvard Ext.
OPTION B
PLAN B — BC retake for A+ + AP Stat + AP Lit + Mentored research continuation + Internship period
Safer; neutral signal
+0.4 pp reach avg
+Lower stress, preserves GPA
+Research time protected
No new depth signal past 11th grade
Reach readers may see plateau
Weakest narrative of the three
OPTION C
PLAN C — Multivariable (LHS) + AP Lit + AP Stat + Independent study (ML) + Competition calendar (USACO Platinum, Putnam)
Maximizes distinctive signal
+2.6 pp reach avg
+Competition results shift narrative from ‘strong student’ to ‘exceptional’
+Independent study produces tangible artifact
Putnam is noisy — no medal is possible
No dual-enroll transcript to show
§ 8
Application concerns

Where college admission readers will get skeptical — before they do.

Every application has weak points. The model flags the top five, ranked by severity, with the concrete rewrite or action that neutralizes each one. This section is the one parents print out.

§ 8
HIGH
#01
Non-profit leadership lacks measurable outcomes
‘Code for Lexington’ is described as ‘serving 147 students’ across the activities list — but no outcome metric (completion rate, placements, follow-up learning) appears anywhere. Readers discount unmeasured impact at a known rate.
RecAdd two outcome data points to the activities list and Common App description. Draft language in §9.
HIGH
#02
Activity list concentrates signal in one domain
Seven of twelve activities cluster around CS/math competitions and research. Readers at humanities-rich institutions (Stanford, CMU Dietrich-adjacent reads) pattern-match this as narrow. Piano Grade 8 currently sits at the bottom of the activity list where it reads as filler.
RecReorder activities so piano and non-profit surface in slots 4–5. Re-weight hours to reflect actual commitment, not perceived prestige.
MED
#03
Narrative draft reads STEM-only
The personal narrative currently foregrounds the USACO arc and the protein-folding project — strong, but both sit inside the same technical frame. At humanities-rich institutions readers look for a second register of thought, and the Code for Lexington work is the most obvious lever not yet activated on the page.
RecRework the middle third of the draft to ground the non-profit work in a specific student outcome. See §9 for three framings scored against each target school.
MED
#04
Course rigor reads flat in senior year
Current senior schedule (AP Stat, AP Lit, CS elective) shows no post-calculus math and no dual enrollment. For a reach-heavy CS list, readers expect to see either multivariable/linear algebra or an independent-study artifact by the time the mid-year report lands.
RecSee § 7 — Plan A (dual enroll Harvard Ext.) or Plan C (LHS multivariable + ML independent study) both close this gap.
LOW
#05
Target list leans heavily reach
Eight of ten listed schools fall into the reach band at Arjun's current profile. That's not wrong for a CS applicant with this record, but the likely band is thin enough that a single soft ED outcome leaves the spring narrow. Two of the target-band schools we suggest in §11 would restore balance without trading down.
RecAdd one target and one likely from §11's additional school recommendations. Keep the reach list intact.
§ 10
Schools you haven't considered

Five suggestions the family list is missing — and why.

Most lists are built from reputation rankings and peer behavior. The report surfaces schools where this specific studentfits, including ones outside the family’s geographic or cultural radius. A fit-weighted list, not a prestige-weighted one.

§ 10
Harvey Mudd
Reach
18%Projected acceptance
Small, intense, proof-heavy CS; aligns with Arjun's competition-math bent. Core curriculum matches his comfort zone better than most reaches; writing-sample requirement plays to his above-average verbal score.
Why not on list:Missing because profile prioritized CS rankings, not fit.
Rice (CS, Residential College)
Target+
34%Projected acceptance
Under-applied by Northeast STEM students. Residential-college culture rewards the community-building signal in Code-for-Lexington. Houston tech ecosystem growing.
Why not on list:Geographic bias. Rice doesn't surface in most Lexington HS peer lists.
Vanderbilt (Blair/Engineering double)
Target
42%Projected acceptance
Only works if he'd consider continuing piano formally — your conversation notes suggest yes. Vanderbilt reads this pairing as genuinely distinctive; acceptance model shows a 7 pp lift for applicants with Grade 8+ performance audition track.
Why not on list:Requires a secondary interest the family doesn't initially surface as ‘admissions material.’
University of Waterloo CS/Co-op
Target
56%Projected acceptance
Best-in-class co-op pipeline for a student who wants real-world CS experience pre-grad. Canada tuition + F-1 visa math works. USACO Platinum translates directly into Waterloo's CCC/CCO-oriented admit lens.
Why not on list:Outside US conventional list; few LHS families consider it.
Williams (CS + Math)
Reach
16%Projected acceptance
Higher admit rate than comparable reaches, smaller CS department but exceptional student-faculty ratio. The research mentor signal (Harvard BioChem) is fungible there in a way it isn't at MIT. Consider if liberal arts appeals at all.
Why not on list:Liberal arts not yet on radar — worth a conversation.
INPUT · YOUR STUDENT'S DATA
Comprehensive profile
Transcript (4 yrs)
AP / IB scores
SAT / ACT attempts
Senior coursework
High school catalog
Activities
Summer programs
Personal narrative
Target school list
Major & preferences
OUTPUT · STRATEGY
Strategic report
Executive summary
Profile snapshot
Strengths read
Risk flags
Per-school probabilities
Test-score deep dive
Senior coursework
Extracurricular assessment
Narrative read
ED / EA strategy
Added school recommendations
Action plan
Why College Signal, not a private consultant

More depth than any human could produce. In twelve minutes, not three months.

THIS IS US
College Signal
ALT 1
Private consultant
ALT 2
School counselor
ALT 3
DIY / Google
Personalized to student's data
Acceptance probabilities
Per-school, modeled
Anecdotal
Hours of advisor time / student
AI — none required
40–80 hours
3–6 hours
n/a
Time to first deliverable
Minutes, not months
4–8 weeks
1 month
n/a
Sample size behind model
Deterministic engine + catalog of 1,500+ US colleges
Advisor’s memory
Prior class
Google
Cost
$249 one-time
$5,000 – $25,000
Free
Free
Inside the report

Eleven sections plus an executive summary. One personalized read.

Every section is generated for your student specifically— no templates, no generic advice.

View sample report
§ 01
Acceptance Probabilities
Master table. Color-coded. Per school, per round.
§ 02
Student Profile
Academic + extracurricular snapshot with context.
§ 03
Strengths
What makes this applicant stand out — and to whom.
§ 04
Concerns
Honest risks and what to mitigate before submission.
§ 05
Test Score Deep Dive
Superscore policies. Retake value calculation.
§ 06
Senior Year Coursework
Course-by-course evaluation with upgrades.
§ 07
Extracurricular Assessment
Tier-scored against national benchmarks.
§ 08
Narrative Evaluation
Personal statement critique, with line edits.
§ 09
ED/EA Strategy
Binding decisions, modeled. Where ED moves the needle.
§ 10
Additional Recommendations
8–10 schools we’d add to your list, and why.
§ 11
Action Plan & Timeline
Month-by-month through Jan 2027.
Built for four audiences

One engine. Four ways in.

For
Parents

Know what you’re buying before you apply. Remove the guesswork from the most expensive decision of your family’s life.

Order for my student
For
Students

See where you actually stand. Understand which schools you’ll get into — and which you won’t — so you apply with clarity.

Run my profile
For
Advisors

Use College Signal as your intake tool. Give every client a data-rich starting point in minutes, then spend your time on what you do best.

Explore Pro
For
Schools

Offer every senior the same quality of analysis as a $20K consultant. District licensing available.

Talk to us

Twelve minutes
from here to clarity.

Your student’s complete admissions strategy, built from their actual data. One flat fee. No subscription.

Begin the report $249 · one-time · one student, one report