Business & Operations

Look at what your customers keep asking about and figure out where your help center is falling short. Ranks the gaps by how many people hit them and how valuable those customers are, gives you the top three with the actual tickets behind each one, and offers to draft the articles right there.

Tell me which conversation and I write your reply. I pull up the customer's history, read your past messages to match your voice, and draft something that actually addresses what they asked, whether it's a bug, a how-to, or a billing question. I never send it and I never promise a date you haven't approved.

Draft the messages that move customers through your lifecycle. A welcome series that walks new signups to their first win across five touches, a 90/60/30-day renewal sequence grounded in what the account actually achieved, a one-off expansion nudge when usage data shows someone hitting a ceiling, or a save message when a customer signals they want to leave. Every draft references real account data and your actual policies, nothing invented.

Get a quick read on your support inbox. Pick what you need: a morning brief that ranks the 5-10 tickets that actually need you today, an overdue check that flags anything about to miss response-time targets or already past due, or a stale-threads sweep that catches conversations you dropped. I scan, rank, and tell you where to start.

Point me at a customer conversation and I give you the short version: where things stand, what you promised, and what the customer is waiting for. Three bullets instead of re-reading a 20-message thread. Especially useful before you draft a reply on something that's been sitting.

Pull your recent customer replies from your connected inbox, read how you actually write, and distill it into a voice profile that every future draft matches. Catches your greeting style, sentence rhythm, sign-off, favorite phrases, and the corporate fluff you never use. After this runs, every reply, article, and lifecycle message sounds like you wrote it.

Draft the outreach you need: a grounded cold email, a sixty-second cold-call script, a post-call follow-up, an inbound reply, a renewal note, or a churn-save email. Every draft is voice-matched to your sent inbox, anchored in your playbook, and stays in a file until you copy and send.

Draft a one-page proposal grounded in the deal: their verbatim problem statement, your proposed approach, scope (in and out), pricing within your playbook's stance, terms, success metrics, timeline, and next step. Anything outside your pricing non-negotiables gets flagged for your approval, never silently committed.

Build a mutual action plan with the prospect: a shared timeline across procurement, security review, budget approval, and legal - with owners (yours and theirs) and dated milestones. Top three risks and any unknown stakeholder surfaced explicitly so you know what to learn on the next call.

Tell me the basics about your company, your ideal customer, your pricing stance, deal stages, and how you handle objections so I can give you better sales help. I ask a few quick questions and write the playbook every other skill reads first. You only need to do this once, and I keep it updated as things change.

Score what you need scored. Pick the subject: every un-scored lead against your ideal customer profile, a single lead's fit and angle, every open deal on health drivers, or every customer's color. I name the top two drivers per row so nothing is a black-box number.

Pull every open deal from your CRM, classify against the playbook's deal-stage exit criteria into Commit / Best / Pipeline / Omit, roll up annual revenue per bucket, and compare to last week's forecast to flag slippage. Confidence per deal is the minimum of stage progress, qualification completeness, and close-plan completeness - no gut calls.

Prep me for a meeting in the shape that fits: a one-page call prep with the right questions for the deal's weakest qualification pillar, or an account review pack with shipped outcomes, usage trend, risks, and renewal runway. Both pull from your playbook and the deal or customer history - no generic templates.

Plan a new customer's first ninety days: kickoff agenda, success metric locked in their words at kickoff, time-to-value milestones, named champions and blockers, and the first-thirty-day risk list. The anchor every later account review and renewal pulls from.

Run the CRM action you need: a hygiene sweep that diffs dupes and stage mismatches without mutating, a natural-language read-only query, a routing pass that assigns GREEN, nurtures YELLOW, and drops RED, or a follow-up task pushed to your task tool. I never mutate without your per-row approval.

Draft a sharp three-sentence reframe to use on the call (acknowledge → concrete anchor-account example → dated next step) plus a short follow-up email in your voice. I look up the objection in your playbook and recent call insights so the response is grounded, not improvised.

Scan your GREEN customers for expansion signals - usage spikes past tier thresholds, team-size growth, feature-request patterns, new-product adoption - and rank upsell, cross-sell, add-on, and seat-expansion opportunities by annual revenue upside over effort to close. Every row cites the signal so you know why I surfaced it.

Get a read on how your sales motion is actually doing. Pick what you need: a Monday rollup across every domain, a cross-call synthesis with playbook edits, a win-loss pattern read, a deep read on a single discovery call, or a pipeline snapshot with the leakiest stage. Every read ends with a concrete next move, not a dashboard.

Turn a transcript or recording into structured notes: agenda actual versus intended, attendees, pains in their words, decisions, action items split internal versus external, and the next step. I match the call to the right lead, update their dossier, and only sync to your CRM with your nod.

Draft a people document for you, like an offer letter, an onboarding plan, a performance improvement plan (PIP), or a stay conversation script. I draft against your comp bands, voice, leveling framework, and hard nos so it sounds like you wrote it. Drafts only, you ship every one.

Scrape every reactor on a LinkedIn post via Apify with profileScraperMode=main, so each row comes back with a full LinkedIn profile attached: experience history, education, skills, certifications, location, connections count. 5-10x more leads than the comment scrape and far richer data per lead. Phase 1 of the reaction-to-outreach pipeline, also runnable standalone.

Find verified emails for a list of leads via Apollo's bulk match (batches of 10), update the Airtable rows with email + company + title + location, and create Apollo contacts under a named label so the leads show up in your Apollo CRM workflows. Phase 3 of both pipelines, runnable standalone if you have an Airtable table populated by the loader.

Run due-diligence on a vendor before you sign. Pick what you need: a fit evaluation that scores them 1-10 against your rubric with risk tier and recommendation, or a compliance check that verifies their frameworks, names their security leadership, and surfaces public incidents. Every claim cited.

Cut through your inbox or calendar so you know what actually needs you today. Pick what you need: an inbox triage that sorts the last 24 hours into needs-me-today, can-wait, and ignore with a specific verb-plus-object action per thread; or a calendar scan that flags overbooks, missing buffers, focus-block clashes, unprotected VIP slots, and meetings without prep across the next 7 days.