SellLabs · Our applied AI research lab
Signal is everything.
SellLabs is pushing the frontier of AI for B2B sales and marketing in complex, regulated industries. We make models and agents more capable at finding and winning high-value deals by working on open problems in buying-signal detection, agent memory, harness optimization, verification and evals, post-training on real sales and marketing data, and the operating models that keep judgment human.
We publish what we learn. What holds up in the field goes into IntelligentNoise.
What the lab is working on.
Signal science
Which indicators, from regulatory milestones to social posts and hiring, reveal a real buying window, and which just add noise.
Relationship intelligence
Finding credible, permissioned paths into complex buying committees.
Agent memory
How revenue agents learn from decisions and outcomes, and forget well enough to keep running for months.
Human–agent operating models
The approvals, handoffs, and controls that make agents trustworthy in consequential sales.
Revenue experiments
Testing new workflows against qualified meetings, pipeline, cycle time, and win quality.
The name.
SellLabs is a nod to Bell Labs, where my grandfather, Robert Lynn, spent 32 years, where my father, Dan Lynn, started his career, and where information theory gave us the mathematics of separating signal from noise.
My father went on to found RelevantNoise in the early 2000s, one of the first technology platforms to use AI to mine the internet and social media for major brands. Signal and noise runs in the family.
We take the name as a standard: rigorous work on a hard, open problem. Ours is making complex agentic systems reliable in sales and marketing. We do it so the companies doing the hardest, most valuable work, in life sciences and beyond, reach the people who need them.
SellLabs is independent and is not affiliated with or endorsed by Nokia Bell Labs, AT&T, or their successors.