About us
At Pencil, we are building the Agentic OS for marketing. We're moving beyond individual AI tools toward a platform where a marketer can drop in a brief and get a fully produced, brand-safe campaign — without manually stitching agents together or knowing which tools to use. Our mission is to make marketing effective and effortless: AI ads are 10x faster and cheaper to make, and 2x better performing, than making them without AI.
We're looking for a Lead PM to own the supply side of our agent marketplace — the quality, coverage, and reliability of every agent a user can access.
The role
You'll own three interconnected problems that together determine whether Pencil's agent catalogue is good enough to be the foundation everything else is built on.
Hero agent quality. The By Pencil core agents are the standard everything else is measured against. You own their quality — defining what good looks like, building the evaluation frameworks that measure it, and driving the improvement loops that get agents to first-pass right output consistently.
Coverage and breadth. Agents currently skew heavily to creative production. Pencil's strategy requires coverage across Strategy, Insights, and Media. You own the roadmap for expanding that coverage — through internal development and through 3P integrations. You own what gets built, when, and to what standard.
The ecosystem. Internal development alone won't scale fast enough. You own the agent builder — making it easy enough for external developers and partners to create high-quality agents on Pencil — and the revenue share model that incentivises them to do it.
Key responsibilities
Own the By Pencil agent roadmap — quality bar, evaluation frameworks, improvement loops, and the release criteria that determine when an agent is ready for users.
Define and scale evals — build repeatable, automated evaluation pipelines that measure agent performance against real customer briefs. Move quality from subjective to measurable.
Drive 3P integrations — own the pipeline of third-party integrations from prioritisation through to launch. Define what good integration looks like and hold the bar.
Build the ecosystem — own the agent builder experience for 1P and 3P developers. Define what makes it easy to create a high-quality agent and what the revenue share model needs to look like to attract serious partners.
Own coverage strategy — identify the gaps in agent coverage that are losing us users or deals. Build the case for what to build next and in what order.
Work backwards from the customer — before any significant build decision, write the customer problem clearly. The technology decision comes last.
Partner with engineering early — not just to hand over specs. Understand the technical constraints well enough to make good tradeoffs and give engineers clear context on why something matters.
Instrument everything — define the metrics that tell us whether quality and coverage are moving in the right direction. Set baselines, measure outcomes, feed learnings back into the roadmap.
Prototype don’t explain - build working and actionable prototypes alongside the design and engineering team to bring your ideas to life.
Your background
Hands-on AI product experience — you've shipped LLM or agentic products in production. You understand evaluation, reliability, and the gap between a demo that works and an agent that works for 10,000 users.
Marketing or creative tech background — you understand the enterprise marketer's world: brand safety, content at scale, the difference between a tool that's approved and a tool that's actually used.
Strong quality instincts — you've built evaluation frameworks before. You know what a meaningful eval looks like versus a vanity metric. You're not satisfied with 'it seems to work'.
Partnership and ecosystem thinking — you've worked with external developers or partners. You understand what makes an SDK or builder compelling to build on.
Technical credibility — you don't need to write the code but you hold your own in conversations about model selection, prompt architecture, and integration design.
Rigorous and fast — you move quickly but you instrument as you go. You don't ship things you can't measure.
You'll thrive here if...
You find the gap between 'this agent works in a demo' and 'this agent works reliably for every brief a global brand throws at it' genuinely interesting to close.
You believe quality compounds — that a high-quality catalogue today is the foundation for everything the orchestration layer can do tomorrow.
You want to define what 'enterprise-grade AI agent' actually means in practice, not just in a pitch deck.
You write to think, not just to communicate. Crisp briefs and honest post-mortems are part of how you work.
KPIs & Success Measures
Eval pass rate: percentage of By Pencil core agents meeting the quality bar on automated evaluation runs against real customer briefs.
Agent coverage: number of agents live across Strategy, Creative, and Media — and MAU in non-creative personas.
Export quality: average export cost and average performance lift per export across agent-led generations.
Ecosystem growth: number of active 3P integrations live and generating usage. Agent builder adoption by external developers.
Delivery quality: P0/P1 bounce-back rate from QA. P2/P3 polish items cleared within one sprint of GA.
Shared north star: % of platform exports originating from an agent-led workflow (currently 15%).
Benefits
25 days PTO plus public holidays, although we operate a Flexible Time Off scheme
Health insurance / private medical cover
Monthly stipend towards wellness, fitness, and learning and development
Remote — work from anywhere in your home country
Enhanced parental leave policies, whether you become a parent through birth, adoption or surrogacy
Access to our Pencil office in The Shard, London for our UK employees
Flexible working hours
Deep Dive About Pencil
Pencil uses AI to make ads. Our mission is to make marketing effective and effortless. We want to become the default way ads get made — because AI ads are 10x faster and cheaper to make, and 2x better performing, than making them without AI.
We're called Pencil because we believe AI will be a tool for creative people, not a replacement — it may even be as fundamental a tool in the future as the pencil was in the past.
Pencil was founded in 2018 with a team from Google, Facebook and Uber with backing from Sequoia and Entrepreneur First. We were acquired by The Brandtech Group in 2023 to pursue a shared vision of bringing GenAI to the Fortune 500.
https://www.trypencil.com
About The Brandtech Group
The Brandtech Group's mission is to be the best company in the world at helping leading global brands drive growth by connecting content, data and media using technology.
It was founded in June 2015 (as You & Mr Jones) by former Havas Global CEO David Jones, with a simple mission to help brands do their marketing better, faster and cheaper using technology. It was renamed The Brandtech Group in January 2022.
Today it generates more than $1BN in revenue and is the largest global digital content partner for many of the world's biggest brands and companies, often using its unique in-housing model. It works with eight of the world's top 10 global advertisers and 49 of the world's top 100. Clients include Banco Itaú, Danone, Google, Intuit, LVMH, Microsoft, Morgan-Stanley, Netflix, Reckitt, Renault-Nissan, PayPal, TikTok, Uber and Unilever.
The Group is one of the most prominent marketing industry voices on the prediction and movement to AI-driven marketing, including Generative AI. It was named one of the World's Most Innovative Companies 2021 by Fast Company, and by CB Insights as one of the World's Most Valuable Private Unicorns.
https://www.thebrandtechgroup.com