Ingest v1.1 production Score 10Cold
Inbound lead · session open · scored live

You are the lead. This is the pipeline scoring you.

Chetan Yadav built a lead scoring system for a GTM team — PostgreSQL, a logistic regression model, a Slack bot that fires on hot leads. This page is that system, pointed at you. Fill in the enrichment record, tell it which role you're hiring for, and the evidence below re-ranks and rewrites itself for that role.

keep scrolling — the intake is only the first stage
01

Ingest

event received

A lead record was created the moment you opened this page, and it is filling in as you read. Every line is inferred inside your own browser — this is simply what any website can see about you.

lead_idresolving…
first_seenresolving…
sourceresolving…
timezoneresolving…
local_timeresolving…
languagesresolving…
platformresolving…
browserresolving…
screenresolving…
viewportresolving…
pointerresolving…
cpu_coresresolving…
connectionresolving…
appearanceresolving…
days_since_engagementresolving…
statusNEW · UNQUALIFIED
02

Enrich

optional · twenty seconds

In the production pipeline this arrives from a data provider. Here you fill it in yourself, and each field shows the weight it carries.

03

Score

logistic regression

The same shape as the model in the repo: test which signals actually predict conversion in SQL first, then fit a logistic regression so every score can explain itself.

model
logit v4
features
5
intercept
−3.20
threshold
0.70
your z
−2.15

        
04

Match

evidence ranked

The vendor record, ranked against the role you selected. Change the role and the order, the bullets and the emphasis all change with it — same facts, argued for the job in front of them.

6 records · ranked for growth / GTM

Background & credentials
educationB.A. (Hons) Economics — Delhi University, School of Open Learning
MBBS (incomplete) — UPHSD Medical College, Metro Manila
CBSE Senior Secondary, PCMB · NEET AIR 2019: rank <7,000
Self-taught accountancy to 11th/12th standard alongside college and tutoring
certificationsGoogle Digital Marketing & E-Commerce · Microsoft 365 Fundamentals
Operations Management, UPenn (in progress) · HSK-1 Chinese, Peking University (in progress)
languagesEnglish · Hindi · Marathi (compliance signage) · Chinese HSK-1 in progress
locationDelhi, India
05

Route

waiting on threshold

At p ≥ 0.70 the pipeline fires a Slack alert and routes the lead to its owner. You can also ignore the model entirely and just email him — it takes this well.

ownerChetan Yadav — Delhi, India
emailchetan00yadav@gmail.com
githubgithub.com/gtm-cyadav
response_slasame day, realistically

The handoff opens your own mail client with this record already written into the draft. Nothing sends without you.

06

Session

what the page noticed

No analytics collected this. It is simply what the page in front of you has been counting since you arrived.

0:00
on this page
0
records opened
0%
read depth
10
final score

You have not picked a role yet, so the evidence above is still arguing for growth and GTM by default.

Everything here runs in your browser — no analytics, no cookies, no tracking pixels. Nothing you type is sent anywhere unless you press a button that says it will be, and then only the fields listed in the record above. The scoring is real logistic regression with the coefficients printed above, and the signals it uses are the same ones the production model uses: seniority, company size, recency of engagement. That version, the one scoring actual inbound leads and posting to Slack, lives at github.com/gtm-cyadav/gtm-lead-scoring.