Skip to content
ConsultEvo

2026 Zappy Award Monthly Winners: What Articulate and Redis Built

Zapier’s June 16, 2026 article named Rachael Silvano of Articulate and Jeff Hirsch of Redis as its first 2026 Zappy Award monthly winners. Silvano built a community contribution-points and rank system. Hirsch established an enterprise operating model for automation ownership, access, and governance.

The two examples address different problems. Articulate’s reported build synchronizes contribution data with member status. Redis’s reported work determines who may use automation, under what controls, and with what support process. The technical patterns below are proposed adaptations, not reproductions of private customer workflows.

Reliable automation starts with the right record and the right owner: define what each event means, and who is accountable for its effects.

Who were the first 2026 Zappy Award monthly winners?

Builder Organization Problem addressed What Zapier reported
Rachael Silvano Articulate Rewarding contributions in E-Learning Heroes A points and rank system connected to community data
Jeff Hirsch Redis Enabling automation across teams without a common operating model Identity, access, shared ownership, restrictions, and intake

Zapier said it had received 50 submissions before naming the two winners. That is historical information from the June 16 article, not a current award calendar. Current program details should be checked on the Zappy Awards page and in the official 2026 rules.

How Articulate’s community rewards system was reported to work

Zapier reported: E-Learning Heroes had 140,000 active members and did not provide the points-economy capability Articulate wanted. Silvano built a supplementary system. The article described a seven-step workflow connecting Airtable with the community platform. JavaScript transformed data and threshold logic selected a result, followed by three webhook calls for authentication, retrieving a user ID, and applying a rank.

Zapier said Silvano built the initial system in an afternoon and used Copilot while working through unfamiliar API calls. That describes a build-time aid, not an AI system making point or rank decisions at runtime. Zapier also reported that the program was live across the community and that peer answers helped offset support costs, but it published no cost calculation, workflow export, API specification, or measured savings.

Public documentation confirms: E-Learning Heroes currently documents a points-based rewards program and qualifying activities in its rewards FAQ. Its rewards guide index names five tiers: Spark, Star, Nebula, Constellation, and Interstellar. That does not verify the historical backend or resolve the article’s reference to six thresholds. The difference may reflect a counting convention, but the public sources do not establish that.

The article’s Airtable link is a general integration overview, not a downloadable version of Silvano’s workflow. Zapier’s Airtable integration documents a polling-based New Record trigger, with frequency dependent on plan. Do not promise an immediate rank change from that trigger. Zapier’s general Airtable integration documentation and workflow overview establish general capability, not the community platform’s endpoint, credentials, permissions, rate limits, or retry behavior.

A reliable design pattern for community contribution points

The following is a proposed adaptation of the reported idea. Start by separating the data into distinct row grains:

  • Contribution ledger: one award, reversal, or manual adjustment per qualifying event.
  • Member summary: one current balance per member and rewards-program version.
  • Rank transition: one attempted or completed rank change, including its result.
  • Audit record: one platform API attempt with the previous rank, requested rank, response status, and timestamp.

A proposed event key is community_id + member_id + source_event_id + event_revision. The revision distinguishes a legitimate correction from a replay. If concurrent runs can process the same event, enforce uniqueness in the system holding the ledger with a database constraint and use a transactional upsert where supported. A find-then-create sequence alone can race and create duplicate awards.

Illustrative contract: these fields are proposed implementation choices, not fields published by Zapier or Articulate.

{
  "input": {
    "community_id": "community_01",
    "member_id": "member_742",
    "source_event_id": "answer_9182",
    "event_revision": 1,
    "event_type": "accepted_answer",
    "points_delta": 10,
    "rewards_program_version": "2026_v1"
  },
  "output": {
    "ledger_version": 184,
    "member_id": "member_742",
    "points_balance": 510,
    "calculated_tier_id": "tier_03",
    "decision": "update_rank",
    "idempotency_key": "community_01:member_742:answer_9182:1:2026_v1"
  }
}

Keep fixed rules deterministic. Code, a threshold table, filters, or paths can calculate the balance and select a tier. AI is unnecessary for arithmetic, exact duplicate checks, or choosing a rank from an approved mapping. If AI classifies an ambiguous contribution type, restrict it to approved labels and require human review before points are awarded.

01Capture the eventRecord the member ID, source event ID, revision, event type, point delta, and program version. Community operations owns the point rules.
02Deduplicate and write the ledgerApply the stable event key and unique constraint. Persist a new award or adjustment before changing a rank, and record whether the event was new or already processed.
03Calculate and validateCalculate the balance and tier from the active program version. Confirm one eligible member, a numeric non-negative balance, a known tier, and a valid destination rank.
04Make and record the updateAfter the platform owner confirms the API and permissions, send the rank change. Store the previous rank, requested rank, response status, and timestamp.
05Route exceptionsSend unknown members, ambiguous matches, invalid mappings, authentication failures, rate limits, and conflicting updates to a named community operations owner.

The community platform’s API endpoint, authentication, permissions, idempotency behavior, and rate limits are not published in the case story. Confirm them with the platform owner before implementation. Retry rank updates only when repeat behavior is known to be safe.

Teams that need help designing or maintaining a governed workflow can review Zapier automation consulting.

How Jeff Hirsch approached automation governance at Redis

Zapier reported that multiple Redis teams wanted to use Zapier without a common owner, success criteria, governance model, or consistent process. Rather than begin with one team’s automation, Hirsch established an operating foundation. The article described SSO, SCIM, and user provisioning across five teams, along with shared folders, service accounts, restrictions on sensitive app actions, and an intake and prioritization process.

Zapier credited Akissi Lewis with project management and Jason Illiscas with the Okta integration. It also reported that Hirsch used MCP and the SDK as part of Redis’s broader AI-tooling effort. The story does not reveal Redis’s internal architecture, available tool scopes, or whether every action was permitted.

Current documentation explains the capability context, not Redis’s historical configuration. Zapier’s SAML guide documents single sign-on. Its SCIM guide says current SCIM provisioning requires an Enterprise account with Workspaces. Current enterprise administration documentation describes app and action restrictions, approvals, shared connections, alerts, analytics, audit logs, and ownership controls. Those are current options, not controls the article confirms Redis used in the historical rollout.

Demand is not adoption

Zapier reported that teams requested access and proposed use cases before formal enablement. That shows interest, not sustained use or business impact. Measure activated builders, published workflows, recurring runs, successful-run rate, retained use, outcomes, and support burden.

A practical rollout sequence for governed enterprise automation

  1. Assign ownership and intake. Record a business owner, technical owner, use case, systems, data classification, read and write needs, expected outcome, support contact, and retirement owner. Requests without an accountable owner remain in review.
  2. Configure identity and workspaces. Set up applicable SAML and SCIM controls, then test joiner, role-change, deactivation, reactivation, and offboarding behavior before broad access. Confirm current Enterprise and Workspace prerequisites for SCIM.
  3. Set action boundaries. Define approved apps and actions, service-account ownership, and human review for sensitive workflows. Apply the same policy to MCP and SDK access. An AI interface does not replace permissions.
  4. Test operations. Verify failure notifications, recovery, rollback, credential ownership, and support escalation. Test expired connections, API limits, disabled users, and failed identity provisioning.
  5. Measure and retire safely. Compare outcomes with a baseline using recurring runs, successful-run rate, business results, and support burden. Transfer ownership or retire workflows without an active owner.
Review before rollout
  • A named business owner, technical owner, and support path are recorded.
  • Allowed apps, actions, data classification, and service-account owner are defined.
  • Failure notification, recovery, rollback, and any human approval are tested.
  • A baseline, measurable outcome, and retirement or transfer owner are documented.

For bounded AI access to approved actions, see AI agent implementation. Current Zapier documentation describes MCP as an action interface for compatible clients and identifies the SDK as open beta. MCP requires a client supporting Streamable HTTP. Scope tools and actions explicitly, and verify current terms before deployment.

What the two winning builds have in common

Initiative Source or trigger Control that matters Evidence of progress
Community rewards Contribution or point adjustment Unique event, valid member, deterministic tier Successful rank updates and resolved exceptions
Enterprise enablement Team request or identity event Named owner, scoped access, support process Recurring successful runs and measured outcomes

The examples solve different problems. One synchronizes earned member status; the other establishes how teams may use automation. In both cases, workflow count is a weak maturity measure. A dependable program has an accountable owner, a defined decision rule, safe permissions, exception handling, and an outcome measure.

A note on the award timeline

The June 16, 2026 article said Zapier had received 50 submissions and expected additional monthly awards. Current award pages say submissions are open year-round, while the current rules and marketing page differ on monthly recipient counts. Treat the official current rules as controlling for eligibility and prize terms, and do not present the historical summer schedule as current.