Sales materials & outreach research
Pitch decks, one-pagers, cohort offers, and a living spreadsheet of school and district leads.
Leads + collateral stay current without a sales hire
EdTech · Founder ops
Sales materials, schedules, parent mail, curriculum, test prep, and reports - without four extra roles. Onboarding admin cut from ~6 hours to ~30 minutes.

Acceleo Matemática teaches Grade 6 math through live, teacher-led sessions. The team uses SundayPyjamas AI Suite agents in the no-code workspace (and in Pumble as a teammate) to run school operations - content, communications, planning, and analysis - so founder hours stay on teaching quality.
6h → 30m
Onboarding admin per cohort cycle
2-3 sets
Full lessons + 20 homework Q&A in ~2 hours
4 roles
Coordinator, creator, marketing, sales research
Always on
Agent tagged in Pumble like a teammate
Operational times reported by Acceleo Matemática. This story covers founder ops on AI Suite agents - not the student-facing product AI (see the companion product case study).
Before AI Suite agents
With AI Suite agents
Live teaching, 90+ lessons, homework, assessments, enrollment, parent communication, marketing, and reporting all landed on one founder-teacher. Every hour on welcome emails or sales research was an hour not spent on lesson quality.
Acceleo Matemática needed capacity without headcount - agents that could draft and research under human direction, with a verification loop the team trusts.
Eight recurring workloads - from sales materials and schedules to curriculum, test prep, parent reports, and platform reviews - run through workspace agents and the Pumble teammate. The founder stays the editor and teacher of record.
Pitch decks, one-pagers, cohort offers, and a living spreadsheet of school and district leads.
Leads + collateral stay current without a sales hire
Term calendars, 90-minute session plans, and homework windows - standardized once, reused every cohort.
Scheduling stops being a weekly scramble
Welcome sequences, parent FAQs, homework protocols, and reminders - drafted once, reused every intake.
Onboarding admin: ~6h → ~30m
Lesson sets matched to prior workspace files: key concepts, word problems, and homework Q&A.
2-3 full lesson packs in ~2 hours, agents in parallel
Cumulative practice, mixed-difficulty drills, and answer keys aligned to the curriculum sequence.
Test prep without a second content author
Homework and assessment signals turned into clear student summaries for reteach and parent talks.
Insight before the next live session
Structured progress narratives grounded in verified curriculum and performance data.
Reports that used to stall after class
Engagement, completion, and cohort health rolled into short operating reviews for the next push.
Founder-level ops dashboard in prose
Roles covered under Acceleo Matemática's direction
No-code agents with school files and history loaded - the team directs, agents draft, humans verify.
Acceleo Matemática runs ops from the SundayPyjamas AI Suite workspace. Agents read prior lesson files and curriculum structure, so new drafts match format and concept sequence without re-briefing every time.
The quality loop is intentional: agents draft → the team reviews and curates in a master spreadsheet → verified content loads to the database (~30 minutes scripted) as the source of truth for long-term progress reporting. Nothing student-facing ships without that review.
Marketing strategy sits in the same place: agents audited existing social presence and built launch plans for the founding cohort - targeting where homeschooling families already discuss pain points.
~30 min
DB load after curation
90+
Lessons in the program
20
Homework Q&A per lesson
Tag it like a colleague. It drafts with company context while the human conversation continues.
Acceleo Matemática added the SundayPyjamas agent to Pumble as a team member. @mention it in-thread; it responds with a link to the full workspace chat, drafts with historical context already loaded, and works in parallel.
Meeting transcripts become decisions and next steps. Ideas fired from mobile between errands get captured and executed instead of dying on loose paper. The mindset shift: an agent is not a tab you open - it's a colleague that's always available.
@agent
Same habit as tagging a human
Mobile
Tasks between appointments
Specialists whose time is eaten by adjacent admin can reclaim hours without hiring a full ops stack. AI Suite agents hold context across files and chat, so each draft gets cheaper than the last.
Governance stays human: draft → review → verify → ship. That loop is the pattern enterprises need before they scale agent workload across teams.
Student-facing product AI (teaching assistant, Curiosity) is a separate deployment on the Apps API - covered in the companion product case study. This page is only about how Acceleo Matemática runs school operations.
“The agent understood the structure of our existing lessons - key concepts, format, sequence - and produced content we could actually use. Writing the welcome email's explanation of the teaching assistant manually would have been genuinely hard; the agent's version was more professional than what we'd have written ourselves.”
“We stopped tagging each other for everything. Now we tag the agent, and the work is moving before the meeting ends. It's a team member that's always available - I even use it from my phone between appointments.”
Wherever an expert is also stuck administering the practice - tutoring, L&D, professional services - workspace agents plus a chat teammate reclaim capacity under a human review loop.
Tutoring & coaching
Curriculum, parent mail, and schedules under one specialist's review.
Corporate L&D
Course packs, learner comms, and completion reports without a content factory.
Professional services
Proposals, research, and client updates drafted in workspace; approved by the partner.
Start with workspace agents this week. Add a chat teammate when the team is ready.
Pick a topic to start, or type your own question below. This page is included automatically with your messages.