Everything you need, built in.
The essentials for teams that want to move fast without giving up control.
Three scores. One number.
Fit: firmographic match to your ideal customer profile. Intent: behavioral signals — pricing page views, competitor research, role searches. Engagement: email opens, meeting attendance, response velocity. Blend them with weights you control.
- Fit: company size, industry, tech stack, title match
- Intent: buying signals from first-party and third-party data
- Engagement: replies, opens, meeting attendance, recency
- Custom: add any field as a signal with a single click
Your scoring model gets smarter every two weeks.
On a 14-day cadence, Atlas ingests new closed-won and closed-lost outcomes and re-weights every signal. If 'VP Revenue' titles started converting 2× better this quarter, the weight for that title goes up — automatically, with audit log.
- 14-day retrain cycle with before/after accuracy diff
- Manual retrain on demand (e.g. after new product launch)
- Audit log for every weight change
- A/B test new weights in shadow mode before promoting
Teams ship revenue with this.
Real-world use cases across every revenue function.
Frequently asked questions
How is this different from a rule-based score?
Rules (A-B-C-D, or +10 for VP, +20 for big company) are static and go stale. Our model uses ML weights tuned against your actual outcomes. And you can override ML weights with rules if you prefer — both models supported.
Can I see exactly why a score is what it is?
Yes — top 5 drivers surfaced on every score. SHAP-based explanations for analysts who want deeper visibility.
Can I run this on existing leads?
Yes — bulk-score your entire database on first setup. Typically takes 4-8 minutes for 100K leads.