Content pipeline
Content pipeline running on autopilot
Keyword research and publishing pipeline for an in-house SEO stack. Indexed pages generated, linked, and maintained without a writer on payroll.
Publishing without an in-house writer
Market intelligence
The BD team reads one email instead of scrolling LinkedIn for three hours
400 target accounts monitored nightly across Companies House, LinkedIn, trade press, and hiring boards. Signals are scored and summarised into a single 7am brief.
Client
UK professional services firm
Build time
5 weeks to production
Headline outcome
Briefing without an analyst
The problem
Consulting and professional services firms pay senior people to spot triggers, a new CFO, a funding round, a hiring spree, a regulatory filing. That signal-spotting happens in 30 browser tabs, most of it never gets captured, and it's the first thing to drop when delivery gets busy. Result: the BD team hears about the right trigger a week after a competitor pitched on it.
The system
A nightly monitoring pipeline across four data sources. Every account has its own signal profile (person changes, financial events, hiring patterns, content themes). Raw events are de-duplicated, scored against a buying-intent rubric, and shipped as a prioritised morning brief with direct pitch angles per account.
How it runs
The workflow, end to end.
01
Source ingestion
Companies House filings API, LinkedIn via authenticated scraping, trade press RSS + scraping, and hiring boards (Indeed, LinkedIn Jobs, client careers pages).
02
Entity resolution
Each event is resolved to a canonical account in the CRM. Duplicate and near-duplicate events across sources collapse into one record.
03
Signal scoring
An LLM grades each event on urgency, buying relevance, and angle strength. Scores tune over time from BD team thumbs-up/down feedback.
04
Morning brief
A ranked digest is generated at 6am: top 10 accounts, the specific signal, suggested opener, and a link to the source. Delivered to Slack + email.
05
CRM write-back
High-signal events are logged against the account in the CRM with the source URL, score, and suggested next action.
Outcomes
What it moves.
400
Accounts monitored
~20
Analyst hours saved / week
~6%
Signal-to-meeting rate
<24h
Time from trigger to outreach
Stack
What it's built on.
- Companies House API
- Playwright (LinkedIn + careers scraping)
- Claude (scoring + brief generation)
- Slack + SendGrid (delivery)
- Postgres
- Trigger.dev (nightly orchestration)
Professional servicesSales intelligenceMarket monitoring
Sell to mid-market? Let's scope the accounts you'd want monitored.
20 minutes. No slides. One process, one prototype.
Book a 20-min introMore builds
View all