The judgment engine for revenue.

Built for agent-first teams.

Dependable decisions, not plausible drafts. drft reads your deal conversations, grounds every decision in your own history, and shows the evidence behind it.

Request early access ↗ See how it works →
drft deal view: the deal's state, the evidence, the recommended next action, and grounded comparable deals
Illustrative data
The gap
No shortage of sales systems. A shortage of judgment.
Every tool feeds in more information. The decision still lands on one overloaded human, and judgment is the one thing that does not scale.
EVERYTHING FLOWS IN ONE HUMAN DECIDES MOST STALLS CRMEmail threadsMeetingsCalendarSlack / TeamsBuyer DMs 1 rep Follow-upsProposalsForecasts stalledgone cold
The shift
drft turns judgment into scalable infrastructure.
Grounded in your own deal history · owned by the company · sharper every deal
Product

A living read on every deal, with the receipts.

drft knows where every deal stands, and can prove it.
Ask drft

Just ask. Skip the dashboard.

Ask in plain language and drft does the work, so you never have to dig through a dashboard to know what to do next.

drft natural-language chat: ask what needs attention and get a grounded answer
Pipeline

Your pipeline reads itself.

The most at-risk deals surface first, each with a next step. No fields to fill.

drft pipeline, deals ranked by risk with a next step on each
The decision

The decision, and the proof.

Open a deal to see its real state, the next move, and the exact messages it stands on.

drft deal view with next action, reasons, evidence citations, and matching won deals
Deal memory

Grounded in your own history.

Every recommendation is backed by the past deals that rhyme with this one, ranked by how closely they match.

drft comparable deals, similar past deals ranked by match
How it works

One engine, from conversation
to decision.

A high-level look at how a message becomes a decision you can trust.
drft // architecturehigh level
Inputs

Conversation

EmailMeetingsCRM
→
Memory

Owned deal memory

DealsPeopleEntity graph
→
drft core

Judgment engine

WeighGroundVerify
→
Output

Decision + receipts

What changedNext moveEvidence
Precedence enginegrounds every decision in the past deals that rhyme with it.
Observereads & structures
Decidewith evidence + precedent
Recommenddrafts the next move
Execute w/ approvalyou click yes
Execute autonomouslyonly where earned

Autonomy is earned, not assumed. Every action stays auditable, and a human approves before anything acts.

Who it's for

For teams scaling faster than judgment can keep up.

Fast-growing, agent-first revenue teams, where deal volume outruns the handful of people who actually know what to do next.
B2B SaaSDeveloper & AI toolsFintech & paymentsData & cloudProfessional servicesMarketplaces
What's next
Your reps use it today.
Your agents run on it tomorrow.

Same engine, same evidence. The judgment your team trusts is the judgment your agents will run on, accountable from day one.

drft answering a question about a deal, grounded in the deal's real state
Team

The team

Arpit Sharma

Founder & CEO

12 years in tech, building ecosystems and business development across the globe. IIM Ahmedabad.

LinkedInXEmail

Vikas Pandey

Co-founder & CTO

Two decades building systems, NLP, and the agentic trust layer behind dependable decisions.

LinkedInEmail
The judgment engine for revenue

Put drft on your own pipeline.

Request early access ↗