We Built Our CRM From Scratch. Here's What 64 Versions Taught Us About Outbound
Sales tools are everywhere. Most are built by people who've never sent a cold message. We built our own — 64 releases later, here's what's inside.
Mar 2, 2026
Sales tools are everywhere. Most of them are built by people who've never sent a cold message in their life.
We're a small team that runs real LinkedIn outbound every day — hundreds of messages, dozens of conversations, actual pipeline. When no tool did what we needed, we built our own. 64 releases later, here's what's inside.
The Problem With Most CRMs
They're built for reporting, not for selling. Reps waste time logging activities. Managers stare at dashboards that show what happened last quarter. Nobody gets help with what to do right now.
We wanted something different: a system that removes friction from outreach and gets smarter the more you use it.
What We Built
AI That Actually Helps You Sell
Not a chatbot bolted onto a sidebar. Our AI reads your pipeline data, researches prospects in the background, and generates personalized messages based on what's working. It auto-classifies every interaction by intent (intro, follow-up, objection handling, meeting set) so you always know where a conversation stands.
There's even a coaching chat that pulls live CRM data mid-conversation — ask it "who should I follow up with today?" and it gives you a ranked list with context.
Messaging Analytics That Close the Feedback Loop
Most teams have no idea which messages work. We track every approach version through the full funnel: sent, opened, responded, accepted. You can A/B test hooks at scale, iterate on phrasing, and see exactly which angles convert for which verticals.
The messaging playbook lives inside the tool. Tone rules, proven hooks, anti-patterns — all editable, all feeding into AI generation.
Zero Manual Logging
A Chrome extension captures LinkedIn activity in real time: sent messages, connection requests, profile visits, full conversation threads. Inbound replies auto-create follow-up todos. Reps never have to alt-tab to update the CRM.
Emails go out from the contact page via SendGrid, with built-in cadence rules (timing, angle changes, soft exits, max 3 follow-ups).
Sales Pulse for Daily Rhythm
Daily touch quotas with a team heatmap. You see who's active and who's going cold — at a glance. Consecutive-day streaks with at-risk badges and Slack alerts keep momentum visible without micromanaging.
A quick daily check-in feeds into streak tracking and the AI coaching engine.
Gamification That Actually Works
Every outreach action earns XP. 25 levels across five tiers — Prospect, Connector, Closer, Rainmaker, Legend. 37 achievements across activity, outcomes, quality, streaks, and hidden surprises. Quarterly leaderboards with weekly movers.
It sounds gimmicky until you see a rep sprint to hit "Rainmaker" before quarter-end.
Pipeline With Real Revenue Modeling
Deals track MRR, one-time fees, and mixed revenue. Stage probabilities give weighted pipeline values. BANT qualification fields feed directly into AI-generated outreach so emails reference the prospect's actual situation, not a template.
Batch Cohort Analytics
Group contacts into campaign batches and track cohort-level conversion. Compare batch performance side by side. See which campaign, which audience, which approach version moves the needle.
Team Scorecard & Multi-User
Each rep sees their own pipeline. Managers see everything: touches, response rates, deals progressing, streaks. Company overlap guards prevent two reps from unknowingly targeting the same account.
Mobile-First, AI-Native
PWA that installs on your phone. Telegram bot for managing your CRM from chat. A JSON API and MCP server so AI assistants can read and write CRM data directly.
Why We're Sharing This
We built this because we needed it — and because we believe the best way to show expertise in AI and sales automation is to use it every day. Every feature exists because we hit a real problem running real outbound.
If you're running LinkedIn outreach and any of this resonates — or if you're curious what a purpose-built AI sales stack looks like in action — reach out. We're happy to give you a look under the hood.
Built with Go, HTMX, PostgreSQL, and a lot of real conversations.