For B2B sales teams

More pipeline, same team.

I build AI systems that help B2B sales teams generate more pipeline without hiring more people. More accounts researched per rep, leads found from live signals, and follow-up within 10 minutes. It starts with one audit of how your team finds and wins deals.

Book a 15-minute intro callSee how it works ↓
Pipeline, live
ClayClay enriched the companyHeadcount, tech stack, contacts✓
ApolloApollo verified the decision makerEmail and direct dial confirmed✓
AttioAttio created the recordStage set, owner assigned✓
SlackSlack pinged the owner"New high-fit lead, ready to work"✓
✓ Deal moved to Won

The stack I build on

Claude OpenAI SupabasePython Attio HubSpot Clay Apollo n8n Stripe Slack
What changes

The work that eats a sales week, done by a system.

Your reps lose hours every week to research, list building and chasing replies. Give that work to a system and the same team works more accounts, reaches buyers sooner and lets fewer deals go cold.

01

More accounts researched per rep, per week

Every account lands in the CRM with a brief already written: the company, the right people, recent news and a reason to call. Your reps get back the hours they spent in browser tabs, walk into every conversation better prepared, and work far more accounts each week.

Northwind AnalyticsResearching
CoFintech, 42 staff, Series A in March
PeHead of Sales, RevOps lead, CFO
NwOpened a London office last week
→Why now: hiring 4 SDRs, no CRM owner
46accounts briefed this week
23 hrsrep research time saved
Example account, illustrative figures
02

Leads found from signals, not lists

Funding rounds, new sales hires and job changes, scored against your ideal customer the day they happen. Your team reaches buyers while they are ready to talk, instead of months later from the same list your competitors bought.

Example signals, scored out of 100
03

Follow-up within 10 minutes

When a prospect replies, a response is drafted and routed to the right rep within 10 minutes, ready to send with one click, or sent automatically for the simple ones. Warm leads stop going cold while they wait in someone's inbox.

with the system
10 minutes
typical
still waiting2 to 3 days
Time from reply to response, illustrative
04

A pipeline you can forecast from

Clean records and reporting that reconciles, so the pipeline number you take to the board is one you can stand behind, and every agent above works from data it can trust.

MayJunJulAugSepOct
Each month built from the months before it, dashed is the forecast, illustrative
How it works

Four stages. Stop after any of them.

Every engagement starts in the same place, because scoping a build off a guess is how projects go wrong. Each stage stands on its own and is priced on its own.

Stage 1

The Pipeline Audit

A roadmap of where deals leak and what each fix is worth. Yours to run yourself, or with me.

£1,000, off the build price
Stage 2

The foundation

Your CRM and data cleaned, connected and trusted.

Fixed price, quoted from the audit
Stage 3

The engine

Research, signals and follow-up agents built and live.

Fixed price, quoted from the audit
Stage 4

Retain us

Tuned, extended and fixed as your team grows.

From £1,000 a month
Start here

The Pipeline Audit

A paid look at how your team finds, works and closes deals: the lead flow, the sales process, and the CRM and data underneath. You get a written blueprint of what to fix first and what each fix is worth.

Lead flow
leaking
Sales process
slow
CRM and data
gaps
AI opportunities
4 found
Fix first Reply routing, then account research
Example audit findings, illustrative
  • Opportunities ranked by pipeline added and hours saved, with team interviews if they would help
  • A roadmap that is yours to keep. Implement it yourself, or work with me on it, no obligation either way
  • The £1,000 comes off the build price in full if you go ahead
The Pipeline Audit, £1,000, redeemable against the build
See what's involved
Then

The build, then retain us

I build the top opportunities from the audit to a fixed scope and a fixed fee, with the limits written down. After that, I stay on month to month to keep the agents tuned and add new automations as the business changes.

Research agent46 briefs this week
Signal feed12 new today
Reply drafting9 drafted today
Request queue1 in progress · 3 done this month
Example system, illustrative
  • Builds scoped from the audit, fixed fee agreed up front
  • Agent tuning and new automations, one live request at a time
  • You buy capacity and a request queue, not a timesheet
Build: fixed price, quoted from the audit
Retain us: from £1,000 a month
What gets built
£

What is manual sales ops costing you?

Three sliders: how many people, hours lost a week to research and CRM admin, and what an hour costs you. No email needed.

Work it out
Proof

Systems in daily use, not slides.

65%
RecruitmentAI agents

A major UK recruitment group

Reduction in manual candidate research time after building an AI sourcing and screening engine that runs the first stage of every placement.

  • 60+ employer communications handled daily by the AI agent
  • Candidate shortlisting automated, humans only at escalation point
  • Live and in daily use across the business
300+
B2B SaaSOutbound

An enterprise B2B software platform

Personalised landing pages deployed for target accounts, each built from live signal data, alongside a signal-fed outbound engine for the BDR team.

  • Account-specific pages generated from funding, hiring and tech-stack signals
  • BDR team freed from manual account research per prospect
  • Outreach personalised at scale without additional headcount
247
Professional servicesOutbound

A B2B events and sponsorship business

Target accounts identified, enriched and enrolled into personalised email and LinkedIn sequences in the first automated run. List building went from a manual half-day to fully automated.

  • Signal-based scoring across funding, hiring and industry fit
  • Decision-maker contacts enriched and verified per account
  • Scored prospects delivered weekly for the length of the engagement
Sonny Driscoll, founder of APEX AI
Why me

An AI and data engineer who started in sales.

Most people who do this came out of sales, or out of engineering. I did both, in that order. Four years on the phones first: a bank, a recruitment desk, then selling holiday properties abroad and sponsorship at a B2B technology events business. At 25 I went back to university and took a first-class Computer Science degree. The decade after that was production data systems at GE, BT, Innovate UK and confused.com, then a solutions engineer seat at Cubed, sitting between the engineers and the sales manager on enterprise deals.

The AI half is not a recent pivot. There are systems of mine in daily use right now: an agent handling the front of a recruitment desk, a public scoring tool that runs four APIs and two models per request, and the assistant answering questions in the corner of this page. All built with real tool access, running against live data, with a defined answer for what happens when a model returns the wrong thing.

Anyone can call a model. Knowing where the cost lands, what to do when it fails, and how it joins a CRM schema without breaking it, that comes from the engineering. Knowing what a rep will actually open on a Monday morning comes from having been one.

Founder-led. I scope and build the work myself, and bring in specialist help when a project needs it. If a scope outgrows that, I tell you rather than bury it.

The longer version →

Worked at
BTGeneral ElectricInnovate UKconfused.comCardiff Metropolitan University
Official Attio Expert Partner★ Upwork Expert-Vetted · top 1% Clay Bootcamp graduate
★★★★★5.0 rating · 100% Job Success Score · Expert-Vetted Top 1%
Questions

Straight answers.

More of the right accounts researched and contacted each week by the reps you already have, new opportunities picked up from live signals rather than bought lists, and replies followed up within 10 minutes. The audit measures where your team is today and puts a number on each of those before anything gets built.

Often, yes. The difference is in the parts nobody demos: what the system does when the model returns the wrong shape, where the cost lands at volume, how it gets access to your real data safely, and how you know it is still working in three months. Ask me on the call what happens when a run fails.

No. The agents, the signal feeds and the follow-up all write into whichever CRM your team already works in, HubSpot and Salesforce included. When a team needs a new CRM, Attio is usually what I build on, and I am an Official Attio Expert Partner. More on that here.

Teams with no sales process yet to improve, anyone wanting AI because it is AI, and anyone after a cold email campaign with nothing behind it. If any of those fit, I will tell you on the call.

Tell me where the sales week goes.

A short call. I will tell you straight whether there is more pipeline to be had from the team you have, and what it would take. No deck, no pitch.

Book a 15-minute intro call

Prefer email? sonny@weareapexai.com