We are a ten-person team scaling globally. That should be a contradiction. It isn't — because roughly half our workforce is made of agents we built ourselves.
One of our implementation partners posted something last week that stopped us mid-scroll. They described building an agent that listens to stream-of-consciousness voice notes on a run and returns a structured prospect report by the time they get home. No form-filling. No CRM ceremony. Just thought captured and converted to pipeline in real time. What they were describing was Luna — one of our core agent team members, used daily by our internal team and our implementation partners alike. That's exactly how we work at Lua. Not as a proof of concept. Not as a demo we show customers. As the actual, everyday operating system of our company.
in
Partner Post · Re-shared by Lua AI
"It started with a small frustration. My brain usually generates ideas while I run or walk. Stopping to type notes breaks the flow..."
"Now when that happens I just open WhatsApp and start talking. Air-hungry half sentences. Random thoughts. The agent picks it up, and by the time the run is over there's a report waiting — list of potential companies, reasons why they might fit, possible use cases, estimated ROI, even a few contacts."
"I'll update the CRM later." Which we all know in most cases means never. Instead you just narrate what happened. The AI extracts the key information, updates the deal, logs the notes, moves it into the right stage, assigns follow-ups and flags if something is missing."
"Once you experience work flowing like this, it's honestly hard to go back to the old way."
That last line. Hard to go back. That's the thing about crossing the threshold — you can't unfeel what frictionless work feels like. We crossed it a while ago.
This is what our agent team actually looks like.
The Lua Thesis:
Building and managing agents is a core job requirement for every role.
At Lua, every hire — sales, marketing, HR, engineering — is expected to understand, configure, and manage agents as part of their regular workflow.
01 / Commercial Team
The agents closing deals while we sleep.
Revenue operations used to require an army of SDRs, RevOps analysts, and proposal writers. Our commercial agent team handles the infrastructure so our humans can focus on the conversations that actually matter.
L
Luna
Pipeline & Relationship Intelligence
Luna is the memory of our commercial team — she tracks every deal, surfaces what's stalled, reminds the right person at the right moment, and drafts proposals, contracts, and follow-up sequences without being asked.
Example Setup
Luna is connected to our CRM, calendar, and email. After any meeting, she detects the event completion, pulls the calendar notes, and proactively drafts a follow-up email and updates the deal stage. If a deal goes silent for 5 days, she pings the owner with a suggested re-engagement message. If a proposal goes unsigned for 7 days, she flags it with a recommended action.
N
Nova
Lead Enrichment & Opportunity Intelligence
Nova turns a company name into a full briefing — pulling firmographics, tech stack signals, hiring trends, funding history, and competitor context — then scores and slots the opportunity into our pipeline automatically.
Example Setup
When an inbound lead submits a form, Nova triggers immediately — she cross-references LinkedIn, Crunchbase, and job boards to build an enrichment profile, scores the account against our ICP rubric, and prepares a one-page brief for the AE before the discovery call is even booked. The AE walks in knowing more than they would after an hour of manual research.
B
Bob (the Builder)
Implementation Architect & Deployment Planner
Bob takes a scoped deal and converts it into a technical implementation plan — mapping requirements to architecture, generating deployment roadmaps, and flagging integration risks before the contract is signed.
Example Setup
Once a deal reaches the technical scoping stage, Bob ingests the customer's tech stack, the use cases from the discovery call, and our integration catalogue. He outputs a structured deployment roadmap with phasing, estimated effort, dependencies, and risk flags — ready to be reviewed by our solutions team in minutes rather than days.
02 / Marketing Team
Winning search — for humans and for AI.
Modern marketing requires presence in two places: traditional search engines and the AI systems that millions now use to get recommendations. Our marketing agents cover both fronts, and they compound on each other.
Our core marketing thesis: every piece of content we publish should be optimised not just for Google, but for the language models that are increasingly the first place buyers go for vendor recommendations. We call this the SEO/GEO stack — Search Engine Optimisation combined with Generative Engine Optimisation.
M
Malcolm
SEO/GEO Funnel Manager
Malcolm owns the top of our funnel — identifying high-intent keyword clusters, mapping content gaps against competitors, and monitoring where Lua AI appears (or doesn't) in AI-generated answers across ChatGPT, Perplexity, and Claude.
Example Setup
Malcolm runs weekly audits across AI answer engines, querying prompts like "what's the best AI agent platform for B2B sales?" and logging where we appear, where competitors dominate, and what citations those systems favour. He feeds findings into a content brief queue so our team is always writing to close the gaps that matter most for buyer discoverability.
P
Prism
Content Strategist & Brief Generator
Prism turns keyword opportunities and customer conversations into structured content briefs — identifying the angle, the supporting evidence, the internal links, and the structural signals that make content authoritative to both search crawlers and LLM training pipelines.
Example Setup
When Malcolm flags a content gap, Prism receives the keyword cluster and pulls the top 10 ranking pages, analyses their structure and entity coverage, and generates a content brief that specifies: target word count, must-include entities and definitions, internal linking opportunities, and the key differentiating angle Lua AI can own. Our writers go from blank page to structured brief in seconds.
E
Echo
Distribution & Amplification Agent
Echo takes published content and systematically distributes it across channels — adapting the core message into LinkedIn posts, newsletter snippets, and community responses — while tracking which formats generate the engagement signals that feed back into GEO rankings.
Example Setup
When a new blog post is published, Echo triggers automatically. She generates three LinkedIn post variants at different angles (founder insight, customer outcome, contrarian take), a newsletter paragraph, and five community response drafts for relevant forums. She tracks engagement on each variant and reports back which angle resonated — informing the framing strategy for the next content cycle.
C
Cipher
Competitive Intelligence Monitor
Cipher tracks what competitors are saying, publishing, and getting cited for — so our positioning stays sharp and our content team is never caught flat-footed by a competitor narrative gaining traction.
Example Setup
Cipher monitors competitor websites, LinkedIn activity, G2 reviews, and AI citation patterns on a daily cadence. Each Monday morning the team gets a briefing: new competitor content, messaging shifts, customer complaints surfacing in reviews, and any instances where a competitor's framing is being picked up by AI answer engines. We respond with content within the same week.
03 / People Team
The humans who find, grow, and look after humans.
HR is often the last function to adopt new tooling. At Lua, it was one of the first — because the quality of our people decisions compounds faster than any other investment we make.
H
Hellen
Interviewing & Hiring Process Agent
Hellen runs the full hiring process — screening applications, conducting first-round interviews, generating role-specific question sets, collecting structured feedback, and producing hiring recommendation summaries so the final decision is based on signal, not gut.
Example Setup
When a candidate submits an application, Hellen scores it against the role rubric and schedules the first-round interview — which she runs herself, asking a tailored question set based on the candidate's background and the role requirements. She scores responses in real time, flags standout moments or concerns, and produces a structured recommendation before the candidate has even left the call. By the time the hiring committee meets, Hellen has already done the first filter — humans only step in from round two.
F
Friday
Onboarding Experience Agent
Friday ensures every new hire hits the ground running — delivering a personalised onboarding plan, sequencing introductions, surfacing the right documentation at the right time, and checking in at day 7, 30, and 90 to flag anything falling through the cracks.
Example Setup
The moment an offer is signed, Friday generates a personalised onboarding checklist tied to the hire's role, starts a daily briefing sequence that arrives each morning of their first two weeks, and schedules automated check-ins. At day 30, Friday sends the new hire a short pulse survey, synthesises their feedback, and flags any onboarding gaps to the People team before they become retention risks.
V
Vera
Performance & Growth Agent
Vera keeps performance conversations continuous rather than annual — prompting regular check-ins, tracking goal progress, surfacing patterns across team sentiment data, and preparing managers with structured briefings before every review cycle.
Example Setup
Vera sends bi-weekly nudges to team members to log wins, blockers, and learning notes — a 2-minute async ritual that builds a living performance record. Before quarterly reviews, she synthesises those records into a structured summary for the manager, flags any patterns (repeated blockers, unacknowledged wins, signs of disengagement) and suggests discussion prompts. Reviews become conversations, not paperwork.
R
Remi
Leave & Wellbeing Management Agent
Remi handles the logistics of leave management so no human has to — processing requests, flagging coverage risks, keeping balances accurate, and running anonymous pulse checks to surface wellbeing signals before they become attrition signals.
Example Setup
An employee messages Remi in Slack: "I need to take Thursday and Friday off." Remi checks their leave balance, confirms coverage availability with the relevant team lead, approves the request, updates the calendar, and logs it — all without involving a single admin. Monthly, Remi generates a wellbeing summary for the People lead highlighting leave utilisation patterns, burnout indicators, and any team members who haven't taken time off in over 60 days.
"To work at Lua, you need to be building and managing agents every day — no matter your function. It's not a technical skill. It's a mindset about what your job actually is."
The model is the message.
There's a version of the future where AI is something companies buy and bolt on — a copilot here, an automation there, a productivity tool that saves you 20 minutes a day. That's not the future we're building toward, and it's not the company we are.
The more interesting version — the one we believe in — is a world where the ratio of humans to agents is a deliberate strategic decision. Where the question "how many people do you need for that?" is answered by thinking about both kinds of worker. Where your agent team has names, responsibilities, and accountability just like your human team does.
At Lua, that world isn't coming. It's already here. We know because we live in it. We built it, and then we built it for our customers.
If you're a founder, an operator, or just someone suspicious that your current way of working is about to feel very old-fashioned — we'd love to show you what this looks like in practice.
