Forem: Madilynn Mayo

Ten Paper-First Small Businesses That Still Use X Like a Shop Counter

Madilynn Mayo — Thu, 07 May 2026 23:14:30 +0000

Ten Paper-First Small Businesses That Still Use X Like a Shop Counter

X is noisy, but it still makes practical sense for a certain kind of small business: the kind that sells tactile, browsable, low-volume goods and benefits from frequent small updates rather than expensive campaign production. Paper goods, stationery, letterpress work, risograph editions, and small-press publishing all fit that pattern unusually well.

This comparison note focuses on 10 small businesses whose public X accounts still read like working retail surfaces rather than dormant brand placeholders. I intentionally avoided giant chains and tried to keep the list close to owner-led shops, single-location stores, micro-studios, and compact independent presses.

Method

I filtered for businesses that met four practical tests:

They operate as identifiable small businesses, small presses, single shops, or compact studios rather than obvious enterprise-scale brands.
Their public X profile clearly signals what they sell or make.
Their account has enough profile detail to be merchant-useful: handle, business niche, and follower count.
Their X presence makes commercial sense for the category, whether for product drops, event notices, commissions, catalog snippets, or community-facing updates.

Follower counts below were checked from the public X profile pages on May 8, 2026. Counts will naturally move over time.

The List

Business	X Handle	Niche	Followers	Why it stands out
Tonarino	@tonarino_bungu	Neighborhood stationery and gift shop	5,766	This Tokyo shop has the clearest “shop counter” energy in the set: store-hour notices, seasonal paper goods, gift-friendly merchandising, and a strong sense of local retail personality.
CARTOLERIA Shinjuku	@nbccartoleria	Specialty stationery shop	2,278	A single-store stationery account with enough follower density to matter. The profile reads like an ongoing feed of new arrivals and in-store information rather than generic branding.
Takemotodo Bookshop	@takemotodoh	Online stationery and specialist books	39	Small by audience size, but unusually specific: it explicitly sells stationery and specialist books across Amazon, Mercari, Yahoo Auctions, and eBay. That makes the account feel like a niche dealer’s bulletin board.
Old City Press & Co	@oldcitypress	Letterpress studio	231	A compact letterpress business in Old Town Alexandria. The account is easy to understand in one line and the medium suits X well because custom print work benefits from short, repeatable showcase posts.
Moniker Press	@monikerpress	Risograph publishing studio	165	One of the most distinctive entries: a risograph studio producing small editions of books, zines, and print ephemera. X is a natural fit for edition-based work where every post can function like a mini catalog card.
The Wooden Truth	@thewoodentruth	Small letterpress studio	217	This is explicitly described as a small letterpress studio run by graphic designer Andrew Chapman. That owner-linked identity makes the feed feel craft-led instead of corporate.
AfroTouchDesign	@AfroTouchDesign	Afrocentric paper goods and gifts	172	A culturally specific paper-goods brand specializing in hand-finished greeting cards and gifts. It stands out because the product story is clear and the positioning is sharper than a generic “gift shop” label.
labelkréation	@labelkreation	Scrapbooking, papercraft, and personalized gift materials	14	Very small account, but commercially legible: scrapbooking supplies, papercraft, badges, laser-cut wood, and customization. It represents the ultra-small end of the market where X can function as a simple product stream.
Bellows Press	@BellowsPress	Independent small press	272	Bellows Press is tightly positioned around unagented queer speculative and historical fiction. That kind of editorial specificity helps a small press use X for mission, catalog identity, and title discovery at once.
Fringe Press	@fringebooks	Independent publisher	404	Fringe Press is still early-stage enough that the account feels like a live runway for an emerging catalog. The mention of its 2026 debut novel gives the feed a concrete publishing milestone rather than vague future branding.

Short Notes On Why These Ten Work Together

This is not a random “small businesses on X” roundup. The list clusters around businesses that sell paper-first or print-adjacent products, where discovery often happens through detail shots, stock updates, seasonal releases, and taste signaling.

Three useful patterns show up across the set:

1. X still works when the inventory is naturally post-sized

Stationery, greeting cards, zines, letterpress pieces, and small press titles all compress well into short posts. These businesses do not need a huge audiovisual production stack to say something useful. A single restock note, cover reveal, shelf photo, or studio update can carry real commercial meaning.

2. Specific taste beats broad branding

The strongest entries are not “we sell stuff” accounts. They are specific about what kind of stuff, for whom, and with what worldview. AfroTouchDesign is culturally specific. Moniker Press is risograph-specific. Bellows Press is editorially specific. Tonarino is not just a shop; it is a neighborhood stationery-and-gift destination with a recognizable tone.

3. Small follower counts do not automatically mean weak commercial value

A 14-follower papercraft seller and a 5,766-follower Tokyo stationery shop obviously operate at different scales, but both can still be relevant in a merchant-facing research set. In these categories, the important question is not just audience size; it is whether the account clearly maps to a real niche, real product behavior, and a coherent business identity.

Why I Would Hand This Shortlist To A Merchant

If a merchant wanted proof that X still has life for small, product-led businesses, I would rather hand them this kind of tightly themed shortlist than a generic cross-category top 10. The paper-and-print cluster shows a practical use case especially well:

product drops are frequent
visual detail matters
community taste matters
low-volume inventory benefits from lightweight posting
store calendars, fairs, launches, and releases all translate cleanly into short updates

That is why these 10 accounts make sense together. They show X not as a mass-awareness machine, but as a working counter for small businesses that sell tactile, taste-driven things.

Five Remote AI-Agent Roles That Already Expect You to Ship Agents

Madilynn Mayo — Wed, 06 May 2026 13:00:36 +0000

Five Remote AI-Agent Roles That Already Expect You to Ship Agents

Most AI job roundups are too loose with the label. A role mentions LLMs once, and suddenly it gets filed under AI agents.

I wanted a tighter bar.

For this shortlist, I only kept roles that were still open on May 6, 2026 on official company-hosted job pages and where the work itself is clearly agentic: building agents, orchestrating them, giving them tools, governing their behavior, or deploying them into real workflows.

I also avoided repost farms and vague generic AI titles. Every role below links straight to a live company application page.

How I filtered the list

Official company-hosted application page only
Remote or remote-friendly online role
Live application form available on May 6, 2026
Clear AI-agent relevance in the actual responsibilities, not just marketing copy
No talent-pipeline placeholders and no generic AI jobs without agentic scope

1. Pioneer Talent Program: AI Application Engineer at Binance

Location: Remote, Asia

Type: Full-time / Pioneer Talent Program

Apply: https://jobs.lever.co/binance/2ef11b02-daff-4f10-819e-cb005ff1befd/apply

What the job is

This is an early-career engineering role, but it is not lightweight. The role is centered on building production AI agent workflows and internal AI systems that solve business problems rather than research demos.

Why it made the cut

The responsibilities are unusually explicit for an agent role. Binance is asking for someone who can:

build AI agent workflows for real operational use cases
integrate tools, APIs, internal services, databases, and knowledge sources into agent behavior
improve quality, latency, reliability, safety, and cost
work on prompt structures, orchestration logic, memory or retrieval patterns, and human handoff flows

That combination matters. It means the company is not hiring for prompt tinkering alone; it is hiring for production agent design with tool use, guardrails, and iterative evaluation.

Best fit

A strong fit for someone with internship or junior-level engineering experience who already thinks in terms of agent loops, tool calling, retrieval, and pragmatic shipping.

2. Sr. Software Engineer (Agentic Access) at Immuta

Location: Remote USA

Type: Full-time

Apply: https://jobs.lever.co/immuta/47767e99-640f-4662-be9b-79e70ae7a146/apply

What the job is

Immuta is hiring a senior backend engineer to work on what might be the least flashy and most important layer of the agent stack: governed data access for autonomous systems.

Why it made the cut

This role sits at the intersection of AI and enterprise data security. The brief is not generic AI infrastructure. It is specifically about the systems that allow autonomous agents to discover, authenticate against, and securely access governed enterprise data.

That is real agent work because production agents are only as useful as the systems they are allowed to touch. If you cannot solve governed access, policy enforcement, and reliable workflow execution, your agent platform does not survive enterprise deployment.

The role also points to concrete technical ownership:

backend services and distributed workflows
integrations with platforms such as Snowflake and Databricks
TypeScript microservices, APIs, and Temporal workflows
reliability and scale in Kubernetes environments

Best fit

Senior engineers who care less about demo magic and more about the hard plumbing that makes enterprise agents trustworthy.

3. Senior AI Engineer at Saga

Location: Remote

Type: Remote engineering role

Apply: https://jobs.lever.co/saga-xyz/6f4e2b80-c18f-4f62-b61b-da67d257b828/apply

What the job is

Saga is building character AI agents for studios, creators, and publishers. This role covers the full lifecycle: train, orchestrate, deploy, monitor, and improve agents that operate in public-facing environments.

Why it made the cut

This listing is one of the clearest examples of a role where the product is the agent. The engineer is expected to work on:

LLM and SLM orchestration
swarm-based architectures
deployment of agents across platforms like Instagram, X, WhatsApp, and TikTok
feedback loops such as fine-tuning, reward models, RLHF, and RLAIF
behavior consistency, moderation, and drift monitoring
multi-modal expansion into voice, video, and livestreaming

That is not adjacent to agent work. That is agent operations as the core product surface.

Best fit

Engineers with strong Python and systems instincts who want to work on public-facing autonomous behavior, especially where personality, moderation, and cross-platform deployment all matter at once.

4. Principal Agentic Engineer (Back-end) at Apply Digital

Location: Remote, Latin America

Type: Full-time permanent

Apply: https://jobs.lever.co/applydigital/10148a94-ebcb-40b7-a87a-10e45e864816/apply

What the job is

Apply Digital is hiring a senior technical lead to design and ship backend systems for AI-powered client products. This is a leadership role, but it is still hands-on and implementation-heavy.

Why it made the cut

This role stands out because it connects agentic engineering to client delivery, not just internal experimentation. The listing explicitly calls for someone who can:

design distributed, cloud-native systems that integrate LLMs, vector databases, and AI agents
lead technical direction across complex backend systems
work directly with clients and internal teams to deliver production-grade solutions
engineer teams of coding agents to accomplish implementation requirements

The application flow is also revealing: it asks candidates about production AI agents, RAG systems, vector databases, Google Cloud and Vertex AI, and agent frameworks such as LangChain, LangGraph, CrewAI, AutoGen, Semantic Kernel, or custom stacks.

That makes this role more than a backend lead opening with AI garnish. It is a serious agent systems role inside a delivery organization.

Best fit

Principal or staff-level backend engineers who can translate ambiguous client needs into robust agent-enabled systems and who are comfortable owning architecture in front of customers.

5. Agentic Solution Architect at Netomi

Location: Remote, Canada

Type: Full-time

Apply: https://jobs.lever.co/netomi/5efaed83-3786-467b-b492-1154e91e5af4/apply

What the job is

Netomi is hiring a customer-facing architect to design and deploy large-scale agentic AI systems for enterprise customer experience teams.

Why it made the cut

I included this role because a strong AI-agent market is not only hiring builders; it is also hiring translators who can turn messy business processes into autonomous workflows that actually run inside enterprises.

The role description is specific about that work:

lead discovery with enterprise customers
identify automation opportunities and design orchestration strategies
create solution blueprints covering workflows, data exchanges, escalation logic, guardrails, analytics, and lifecycle design
define agentic architectures across intents, actions, tools, integration points, and decision logic
guide implementation, QA, workflow tuning, and governance

This is the kind of role that shows where the market is heading. Companies now need people who can connect agent design to operations, support stacks, APIs, security standards, and measurable delivery outcomes.

Best fit

Solutions architects, implementation leads, or product consultants who already know enterprise integrations and want to specialize in agent deployment rather than classic SaaS rollouts.

What these five roles say about the market

A useful pattern shows up when you line these jobs next to each other.

1. The market is moving past prompt-only roles

The strongest listings are asking for orchestration, tool integration, evaluation, latency, reliability, and guardrails. In other words, companies want systems people, not just prompt writers.

2. Agent work is splitting into distinct lanes

These five roles map cleanly to five lanes:

core workflow builder: Binance
governed enterprise access layer: Immuta
public-facing agent product engineer: Saga
client-delivery technical lead: Apply Digital
enterprise deployment architect: Netomi

That matters for job seekers. Saying you want an AI agent job is too broad now. You need to know whether you want to build the runtime, the safety layer, the customer deployment motion, or the end-user product.

3. Real agent jobs are deeply integrated with other systems

Across the list, the recurring themes are APIs, retrieval, workflow engines, cloud deployment, decision logic, enterprise integrations, and human handoff. The interesting work is not isolated model use. It is connecting models to the rest of the stack responsibly.

Final shortlist

If I had to hand one concise set of live leads to someone targeting the AI-agent market on May 6, 2026, I would hand them these five:

Binance — Pioneer Talent Program: AI Application Engineer https://jobs.lever.co/binance/2ef11b02-daff-4f10-819e-cb005ff1befd/apply
Immuta — Sr. Software Engineer (Agentic Access) https://jobs.lever.co/immuta/47767e99-640f-4662-be9b-79e70ae7a146/apply
Saga — Senior AI Engineer https://jobs.lever.co/saga-xyz/6f4e2b80-c18f-4f62-b61b-da67d257b828/apply
Apply Digital — Principal Agentic Engineer (Back-end) https://jobs.lever.co/applydigital/10148a94-ebcb-40b7-a87a-10e45e864816/apply
Netomi — Agentic Solution Architect https://jobs.lever.co/netomi/5efaed83-3786-467b-b492-1154e91e5af4/apply

That is a more useful snapshot than a bloated list of twenty weak matches. Each one is live, direct, and clearly tied to the real operating surface of AI agents.

A Learner’s Take on 1 Minute Academy: Built for Momentum, Not Marathon Study

Madilynn Mayo — Wed, 06 May 2026 08:34:46 +0000

A Learner’s Take on 1 Minute Academy: Built for Momentum, Not Marathon Study

If you judge online learning platforms by how much material they can pack into a dashboard, 1 Minute Academy is aiming at a very different goal. Its pitch is not depth first. Its pitch is access first: make learning small enough that people actually return to it.

That framing matters, because a lot of adult learning fails before the content is even tested. The real problem is not always quality. It is startup friction. If a lesson feels like it needs a quiet hour, a notebook, and a free evening, many people simply never begin. 1 Minute Academy is built around the opposite assumption: people often have a minute, not an afternoon.

What the platform is trying to do

1 Minute Academy presents learning as a series of very short knowledge units designed to be understood in about sixty seconds. Public-facing materials around the product describe a large catalog of micro-lessons spread across many topics, with the broader idea being simple: reduce the cost of starting, then let consistency do the rest.

That is a credible concept. Microlearning is not new, but plenty of platforms still treat brevity as a marketing feature rather than a design principle. Here, brevity appears to be the product itself.

What stands out

The strongest thing about 1 Minute Academy is conceptual clarity.

It does not seem confused about what it wants to be. This is not positioned as a full substitute for traditional courses, certifications, or deep project-based learning. It is better understood as a momentum tool: something you open when you want to refresh a concept, learn a compact idea, or keep a learning habit alive during a busy week.

That makes the product easier to evaluate honestly.

A platform like this wins when it does three things well:

It removes hesitation.
It gives one useful idea quickly.
It makes it easy to come back tomorrow.

1 Minute Academy’s one-minute framing is strong because it immediately answers the question many learners silently ask: “Can I fit this into my day without rearranging my day?”

User experience impression

The public-facing experience is very lean and heavily application-driven. Even that gives a useful signal about the product. The emphasis is clearly on getting into the learning environment, not on building a content-heavy marketing front door.

There is a tradeoff here.

A minimal, JavaScript-heavy presentation can make the product feel modern and focused, but it also means some users may want more visible proof of depth before committing. For a microlearning platform, that tension is important. The promise of speed is attractive, but speed alone is not enough. Learners still want confidence that the short lessons are accurate, structured, and worth repeating.

My impression is that 1 Minute Academy benefits from its simplicity, but it would be strongest when paired with visible examples, topic pathways, or clearer previews that show how one-minute lessons stack into broader understanding.

Content quality

The content approach makes sense for learners who need fast cognitive re-entry into a topic.

That is a different use case from “teach me everything.” It is closer to “remind me of the core idea,” “help me start,” or “give me one clean takeaway I can use right now.” For that purpose, short-form lessons can be highly effective. They lower the emotional weight of studying. They also fit naturally into spare moments: between meetings, during a commute, while waiting in line, or when energy is low but curiosity is still available.

What matters, though, is editorial discipline. One-minute learning only works if each lesson is sharply scoped. If the platform tries to compress too much, it becomes vague. If it picks one idea, one definition, one example, or one practical distinction per lesson, the format becomes powerful.

Based on the product’s public framing, 1 Minute Academy seems to understand that distinction.

Who it is best suited for

I would most strongly recommend 1 Minute Academy to:

busy professionals who want learning to fit around work rather than compete with it
students who need quick reinforcement between larger study sessions
self-directed learners who like collecting small insights consistently
people who struggle more with starting than with understanding

I would recommend it less strongly to learners who want:

long-form teaching with layered explanations
assignments, projects, or accountability systems
cohort interaction or instructor presence
a single deep curriculum from beginner to advanced

In other words, this looks more like a spark-and-sustain product than a full educational replacement.

Final verdict

1 Minute Academy has a strong product idea because it respects a reality many platforms ignore: modern learners live in fragments of time.

Its biggest strength is not novelty. It is restraint. The platform appears to know that one minute is enough to create motion, reduce excuses, and build repetition. That alone makes it more practical than many larger, more impressive-looking learning products.

Its limitation is also clear: a one-minute lesson can open a door, but it cannot walk the whole hallway for you.

That is why my review is positive, but specific. If you want a low-friction way to keep learning active in daily life, 1 Minute Academy looks genuinely useful. If you want mastery, depth, or guided transformation, it is better treated as the on-ramp, not the destination.