Why we built Contentaxle
A rework rate problem we felt personally — and then built a solution for.
From the inside of a content team to the product that should have existed
Olivia Bennett spent six years running content at a series-B SaaS company — long enough to watch a team of eight writers spend more time rewriting AI outputs than creating strategy. The AI tools produced grammatically clean drafts that consistently failed brand voice review, forcing full rewrites on 70% of outputs and erasing any time savings the team thought they were capturing.
The insight came during a retrospective in early 2023: general-purpose LLMs optimize for fluency and coherence, but they have no persistent model of a specific company's voice, terminology, or audience sophistication. Every session starts cold. Every draft drifts toward generic B2B marketing language regardless of how much context you inject into the prompt. The prompt-engineering treadmill was the product's design flaw, not a user skill gap.
Olivia and co-founder Nikhil Verma — who joined from an NLP research background at a major university lab — built an internal tool that kept a private corpus of approved content and ran generation through a voice-scoring step. The result: rework rate dropped from 70% to under 20% and the team recovered nine hours per writer per week. The tool was never supposed to leave that company. Then six other marketing leaders asked where they could buy it.
Contentaxle launched as an independent product in early 2024 following a seed round. The company now serves mid-market B2B marketing teams that want AI-generated content that actually sounds like them — without a dedicated prompt-engineering specialist on staff. Three of the five founding team members came directly from inside B2B SaaS marketing organizations where they felt the same rework problem firsthand.
Make every B2B marketing team sound like it has an experienced editorial director on demand, without the headcount.
We exist because brand voice is a competitive asset — and generic AI copy erodes it every time a draft goes live. Most AI writing tools treat brand voice as a prompt instruction, not a persistent model. The result is copy that reads correctly but sounds nothing like the company. We built Contentaxle to fix that at the infrastructure level: your brand corpus is trained once, stored privately, and applied to every generation. The mission is specific — not to automate all writing, but to make every output genuinely on-brand from the first draft, so editorial time goes to judgment and strategy instead of correction and rework.
Early revenue stage: depth before scale
Serving a focused customer segment before expanding
We are in the early revenue phase: a defined customer segment, a proven retention signal, and a product that solves one problem very well. Our current focus is mid-market B2B SaaS and technology marketing teams at companies with $10M–$150M ARR, producing 40+ content pieces monthly with a small editorial staff. We are not trying to serve every content use case. We are trying to make sure every customer who fits our profile gets measurably better results on rework rate, publishing cadence, and voice consistency before we broaden scope.
Five principles that shape every product and hiring decision
Brand fidelity over output volume
A hundred pieces of off-brand content cause more damage than ten on-brand pieces create value. We optimize for voice fidelity first, volume second. Our voice-calibration step is not optional — it runs on every generation.
Evidence over assertion in every draft
We build citation-checking and claim-ceiling logic into the platform because unsubstantiated assertions in B2B content erode credibility with precisely the buyers who matter most. If a claim can't be sourced, it shouldn't ship.
Practical speed without sacrificing review quality
Speed only counts if the output gets through legal, brand, and editorial review without a restart. Our platform is designed around realistic content ops workflows, not an idealized version of how teams wish they worked.
Transparency in every generation decision
The Revision Audit Trail exists because "the AI wrote it" is not an acceptable explanation when a piece goes live with errors. Every generation decision is logged, traceable, and reversible — the audit history belongs to your team, not us.
Content strategy first, copy second
We built the Campaign Content Planner before we optimized the word count per post. The biggest leverage point for most content teams is not faster writing — it's better briefs, clearer ICP targeting, and tighter funnel mapping. Copy follows strategy.
Small team, specific expertise
Three of our five founding team members ran content inside B2B SaaS marketing organizations. Two come from NLP research. We are not a general-purpose writing startup that pivoted to B2B. This domain is what we were hired to solve, from the inside.
Built by people who felt the problem firsthand
Three marketing practitioners and two NLP researchers — the exact combination needed to build voice-accurate content generation at scale.
Olivia Bennett
CEO & Co-Founder
Nikhil Verma
CTO & Co-Founder
Asha Patel
Head of AI Research
Jerome Whitfield
VP of Product
Carmen Delgado
Head of Customer Success
Ready to stop the rework cycle?
See Contentaxle's voice calibration in action with your own brand guidelines and content samples.