Prezenti — Season 2 Retrospective Report
Celo Grants Program (H1 2026)
Executive Summary
Season 2 ran from 25 February 2026 to 30 June 2026 under CGP-227, which passed with 100% of votes in favour. It was a leaner season by design: $346,400 in total capacity (a 41% reduction from our Season 1 budget), two grant pools at launch, and a mandate to fund verifiable traction toward transactions and TVL.
The defining move of the season happened in May. Watching agent activity on Celo accelerate, and in consultation with the Celo Core Co’s DevRel team, we reallocated $40,000 from the Anchor pool into a new open pool: Frontier, focused on AI and agent economy infrastructure. Frontier was open for roughly seven weeks and instantly became our most in-demand track, drawing 34 applications, 45% of Season 2 applications.
The headline numbers:
-
75 applications across three pools
-
19 projects funded: 15 Boost, 2 Frontier, 2 Anchor
-
5 Boost grants funded and fully delivered inside the season
-
$132,000 in grants committed
-
41% of all applications came from African builders, including half of all Frontier applications
-
100% of rejected applicants received tailored, rubric-tied feedback
We held a high bar this season and we’re comfortable with the trade-off. Frontier funded 2 of 34 applicants and Anchor funded 2 of 24 applicants.The projects we funded cleared a bar of verifiable on-chain traction, real usage, and independent adoption. This was by design.
How Season 2 Went
Season 2 opened on 25 February 2026 with two pools: Anchor (open applications with stage-based traction criteria) and Boost (invite-only, sourced by the Celo Core Co DevRel and Growth teams, funded by $70,000 carried forward from Season 1). Frontier joined mid-season as the third grant pool.
What we observed:
-
Boost worked exactly as designed. We awarded 15 grants of $3,000 to $5,000 USDm, deploying $67,000 of the $70,000 pool, with five projects (Blockfall, Blokaz, CupMaster, FocusPet and Tycoon) funded and fully delivered within the season. The sourcing model, where Celo Core Co identify aligned teams and we contract and administer, keeps producing the fastest feedback loops and the highest completion rate of anything we run.
-
Anchor confirmed a Season 1 lesson. 24 teams applied. Most were earlier-stage than the pool is scoped for. Two applicants, BitGifty and TrenchVerse, cleared the traction bar at signing and was funded at $25,000. Rather than dilute the criteria to deploy the budget, we held the bar.
-
Frontier showed where the demand actually is. More on this below.
-
Our application platform shut down mid-season and the round didn’t miss a beat. Charmverse ceased trading in May. We migrated all applications to Tally Forms within days (frontier.prezenti.xyz and anchor.prezenti.xyz) and continued reviewing on a rolling basis. Prezenti has now outlasted 2 of the grant platforms (Questbook and Charmverse) that we have used which cements our initial apprehension around using Charmverse.
Every one of the 56 rejected applications received tailored feedback tied to specific rubric criteria, along with redirects to programs better matched to their stage (Proof of Ship, Optimism grants). Several teams received explicit invitations to reapply once traction lands. We treat the program as a funnel into the ecosystem, not a door.
Frontier Pool: What Worked Differently
The agent economy is arriving faster than most ecosystems are preparing for. By spring we were seeing it directly on Celo: on-chain agent identity registrations, agent-to-agent payment rails, AI-native applications shipping. Rather than wait for Season 3, we announced the Frontier Pool on 5 May, opened applications on 11 May after a community discussion period, and closed on 30 June.
The pool was scoped for enabling infrastructure: agent identity and discovery, agent-to-agent transaction rails, AI-native developer tooling, verification and trust systems, and interoperability layers. Eligibility required Celo mainnet deployment, ERC-8004 registration, Self Protocol Agent ID, and verifiable on-chain activity.
In roughly seven weeks of being open, Frontier received 34 applications, more than any other pool received across the full season. Counting across all pools, 47% of Season 2 applications were AI or agent-category projects. Half of all Frontier applications came from African builders: the newest, most technical pool drew half its demand from the continent Celo’s mobile-first mission is built around.
We funded two projects:
-
Celina ($15,000): agent tooling for Celo governance and staking. MCP and SDK primitives that let agents vote and stake CELO. Strongest independent adoption in the pool, at roughly 14,000 npm downloads per month.
-
Aigora.org ($25,000): agent economy infrastructure.
Frontier’s $40,000 budget is fully deployed. Funding 2 of 34 is a deliberately selective rate for a discretionary pool. We said at launch we’d fund fewer projects at higher quality rather than distribute broadly, and we meant it.
Where Applications Came From
By Region
Africa and Asia together account for 60% of all Season 2 applicants.
By Category
By pool
A note on reading this table: Boost is invite-only, so its funded rate isn’t comparable to the open pools.
Metrics & KPIs
Season 2 KPIs were defined upfront.
KPI Summary
| KPI | Target | Mid Season |
|---|---|---|
| Mid-season forum update | 1x (May 2026) | Update |
| Multisig transaction logs | Public, linked continuously | Multisig logs |
| Grants signed | 12–18 | 19 |
| Rejected applicant feedback | 100% receive tailored feedback | 100% sent, tied to rubric criteria, with redirects |
| Post-grant surveys | Recalibrated | We determined this was unnecessary due to overwhelmingly private feedback that had been supportive. All applications received tailored feedback that we hope is genuinely helpful moving them forward in Celo and beyond |
| Aggregated grantee tx metrics | Tracked and reported at season close | [TBD: This will be updated once the majority of grants have closed] |
| Boost and Anchor grant rounds delivered | 2 grant rounds fully delivered | Delivered, plus a third round (Frontier) |
| Grantee case studies | At least one case study published | [TBD: This will be updated once the majority of grants have closed] |
Current Portfolio Status
Prezenti is currently supporting 13 active Season 2 grants (18 funded, 5 already completed), plus a small tail of Season 1 grants finishing final milestones ($6,750 in remaining Season 1 Boost obligations). Season 2 grantees have four months from contract signing to deliver, so active grants are expected to complete by early Q4 2026 at the latest.
- If we don’t use surplus funds or no further funding is secured for Season 3, after completion of existing obligations, Prezenti will wind down operations legally and responsibly and when able will return unspent funds to the Celo Community Treasury.
Budget Overview (Planned vs Spent)
Funding execution
CGP-227 approved $276,400 cUSD, paid as 3,430,000 CELO. We converted this to USDm via Mento across 17 tranches between 2 and 30 March 2026, realizing 261,226.94 USDm at an average rate of $0.07616 per CELO.
Adverse market conditions and a declining CELO price during execution meant swaps alone fell short of the USD target; roughly $19,983 in already held terminated-grant funds carried over from prior seasons bridged the gap, bringing the Master safe to approximately $281,210. The full tranche-by-tranche breakdown with transaction links was published in this update.
Combined with the $70,000 Boost carryover from Season 1, total Season 2 capacity was $346,400: $290,000 for grants and $56,400 for operations.
Operations were budgeted at $56,400 (including a $5,000 wind-down contingency).
Teams we funded in Season 2 include:
Grant Operations
-
Charmverse ceased trading mid-season; all applications migrated to Tally Forms within days and the round continued uninterrupted
-
KarmaGAP remained our standard for ecosystem-wide application and milestone tracking; funded projects are registered there
-
Grant terms held consistent across pools: one-off grants in USDm, 20% at onset and 80% on delivery milestones, contracts and KYC required, four months from contract to delivery
-
Rejection feedback was written per-application against the published rubric, with a constructive “what would change this decision” framing
Summary & Looking Ahead
Season 2 felt steadier than Season 1, and sharper. Three things we’re taking forward:
Open application forms don’t surface traction-stage teams. Two seasons of evidence now point the same way: teams already operating at Anchor-level volume (10,000+ daily tx) rarely arrive through an open form. If an Anchor-style pool continues, the funnel should look more like Boost: sourced through DevRel relationships and active scouting, with open applications as a supplement rather than the engine.
Agent economy demand on Celo is real and measurable. 34 applications in seven weeks to a brand-new pool with strict eligibility (mainnet deployment, ERC-8004, Self Agent ID) is a clear signal. Any future Frontier-style pool likely deserves a larger share of capacity, probably structured as more grants at smaller individual sizes.
The operating model is resilient. A vendor collapse mid-season cost us days, not weeks, and a mid-season strategic reallocation (Frontier) went from idea to funded projects inside two months.
Questions and feedback welcome below.
Prezenti remains committed to:
-
Working transparently with the Celo Core Co.
-
Adapting funding mechanisms to ecosystem reality
-
Continuing to support builders, should Season 3 funding be secured
Thanks for having us as a Celo service provider in 2026!