LeadOS

CONTEXT

Most prospecting tools solve the wrong problem. They optimize for volume, returning long lists of company names with no signal about fit, intent, or timing. LeadOS was built to solve the quality problem with this core hypothesis: a smaller list of highly relevant, enriched contacts with verified details and real buying signals is worth more than a thousand cold names.

THE PROBLEM

Sales teams spend a disproportionate amount of time on prospecting that leads nowhere. Existing tools either return raw data with no context, require manual research to be actionable, or are built for markets where deep local data integration is not possible. LeadOS was designed from the ground up to change that ratio, with the goal being less time finding, more time selling.

PRODUCT ARCHITECTURE

The platform is structured around an AI agent that runs a multi-step enrichment waterfall on every search. The agent reads the user's website and configurates a unique ICP from that, searches for matching companies, scans for buying signals, scrapes company websites, enriches contact data, scores each lead against the ICP, writes personalised outreach drafts, and filters disqualifiers — all in a single automated run. The output is not a list of companies. It is a set of decision-maker contacts with verified emails, fit scores, buying signal tags, and ready-to-send outreach drafts.

TECHNICAL IMPLEMENTATION

LeadOS is a full-stack SaaS product built entirely solo. The frontend is Next.js 14 with a mobile-first responsive interface built in TypeScript and Tailwind CSS. Authentication, database, and row-level security are handled by Supabase with PostgreSQL. The AI agent and enrichment logic run through the Anthropic Claude API. Company discovery uses Serper. Contact enrichment runs through FullEnrich with Hunter.io as fallback. Norwegian company and decision-maker data is pulled directly from brreg.no, which provides a competitive advantage in the Scandinavian market. Email infrastructure runs on Resend. Payments and plan management are handled by Lemonsqueezy with a full webhook integration. The entire system is deployed on Vercel.

WHAT THIS BUILD DEMONSTRATES

LeadOS required solving problems across the full stack simultaneously: designing a product from scratch, architecting a reliable multi-source data pipeline, handling real payments and plan limits, building a mobile-first UI under real constraints, and shipping to production alone. It launched publicly in May 2026 with paying users from day one.