Proposed by: Rich O’Grady (Co-Director, Livepeer Foundation) and Rick Staa (Technical Director, Livepeer Foundation)
Abstract
The Network Engineering SPE is a delegated pool of capital that funds scoped engineering RFPs and retroactive contributor grants, without requiring a standalone on-chain proposal for each initiative. The mission of the SPE is to fund smaller engineering initiatives to make the Livepeer network more observable, performant, and developer-centric. It targets the $2k–$20k range of work that the current SPE process makes too slow to fund. We’ve seen this model work: the Explorer upgrade proved what a vetted contributor with a clear scope and accountable oversight can deliver. This SPE scales that, while exploring a more retroactive model.
Requested duration: 3 month (pilot program) + 2 weeks (retro)
Requested budget: $95,000 USD-equivalent in LPT
The Problem
The Livepeer project needs the ability to execute fast and targeted actions to support users, operators and builders on the network. The current SPE model works well for large, multi-month programmes. It doesn’t work for the $2k–$20k engineering work the network often needs.
The friction is structural:
- A full SPE proposal requires significant time before any work begins
- On-chain vote cycles add weeks of delay — often longer than the work itself
- As AI tooling compresses build cycles, this gap is getting worse, not better
The result: well-supported, clearly-needed work goes unfunded — not because the community doesn’t want it, but because there’s no efficient path to fund it.
This was validated by three public community discussions (March–April 2026) with strong consensus. The Water Cooler on 13 April returned a clear signal: move forward.
The Solution
The aim of this SPE is to build a delegated funding pool with a fast, accountable process which funds critical engineering work fast. This would sit alongside the formal SPE process, which primarily focuses on longer-term (3+ months) initiatives.
This SPE replicates and scales that model across two funding tracks: RFPs dividing up known work and distributing it amongst the community; and retroactive grants funding for work that has already been completed.
Note: Retroactive Grants don’t necessarily go through the Roadmap process.
Eligibility Areas
For the first round of SPE funding, there will be 3 primary focus areas, which are focused on Livepeer’s core stakeholder groups, each with different needs:
Priority 1: [Developers] Developer Portal — The 5-Minute API — Work on the Developer Portal’s critical path: the engineering that takes a developer from docs to first inference call in under five minutes, from any MCP-compatible tool. This includes the Python SDK, BYOC container tooling, payment and auth infrastructure, agentic harness tooling, and any library or scaffolding that removes onboarding friction.
Priority 2: [Delegators] Explorer — Participation & Observability — Work that makes the Explorer a better permissionless participation portal for operators and delegators. This includes staking interfaces, governance participation features, network observability dashboards, and tooling that surfaces real network behaviour to support informed decision-making.
Priority 3: [Orchestrators] Tooling & Infrastructure — Work that helps orchestrators run more reliably, scale capacity, and integrate with the evolving SDK-first architecture. This includes containerisation, orchestration tooling, runtime improvements, go-livepeer contributions, and infrastructure that affects network reliability and operational experience. Note: that this area requires particularly strong community validation before an RFP is promoted
Funding Track 1 — RFPs (~$5k–$20k, 2–8 weeks)
Full proposed process: RFP Process - Network Engineering SPE
Community members post Roadmap Suggestions on the forum, where the suggester drafts initial requirements and deliverables with the community - facilitated by the Livepeer Foundation. After an initial community validation via Discord, the Foundation then publishes a finalised RFP for contributors to apply to, and assigns it to a selected team or contributor within 7 days.
All decisions and disbursements are published with written rationale. Payment is split across three tranches: 25% on signing; 50% on delivery and verification by Review team; and 25% on impact paid at the end of the pilot (in August). Any incomplete work or assessments at the end of the pilot triggers a brief remediation window.
Three Suggestions have already been initially identified and will be scoped with community contributors and then validated by the community to become (or not become) RFPs:
- Roadmap Suggestion 1: Developer Portal & Discoverability platform
- Roadmap Suggestion 2: Improved Explorer as Participation Portal
- Roadmap Suggestion 3: Payment Clearinghouse & Remote Signer
Funding Track 2 — Retroactive Grants (<$5k, retroactive)
Full proposed process: Retro Grants Funding Process - Network Engineering SPE
Build first, then apply. Contributors post a public application on the forum after the work has shipped. The Review Team reviews on a monthly cycle and publishes an approve, partial approve, or decline decision with written rationale. Approved grants are paid in full in a single disbursement.
The primary bar is impact, not completion. Each application must name the problem, who had it, and who is already using the solution. Describe what was built, link to the work, and state the amount requested. Projected usage does not count.
Examples of eligible work and impact:
- OpenRouter integration — 5 developers discovered and integrated with Livepeer via OpenRouter.
- Live video-to-video / BYOC full-stack runner — 3 community members running live video-to-video with BYOC end-to-end — a workflow the network had no viable path for before this.
- Agentic harness tooling — 3 Livepeer solutions cutting time-to-working-agent-loop from days to hours using standardised scripts, tool definitions, and prompts.
Quantifying Impact
Impact means the network becomes more observable, more reliable, or more easy-to-use for Developers, Delegators or Orchestrators. The concrete impact signals that the Review team will look at are all about usage and adoption:
- Named adopter — a specific person or team outside the author is using it in their own work.
- Downstream dependency — another project or contributor is building on top of it.
- Capability confirmed in the wild — someone who wasn’t involved in the build has run it successfully.
It is also expected that the work is maintained and still live by the end of the program. What impact is not: download counts, repo stars, or the builder using their own tool.
What This SPE Isn’t
Not a demand generation fund (coming soon). The Network Engineering SPE funds the engineering substrate: the infrastructure that makes demand solutions like Daydream, Frameworks, Blueclaw, Streamplace possible. Go-to-market, marketing, and end-user services are out of scope.
Not an open-ended bounty pool. Every disbursement requires a verified definition of done. Work must be complete, reviewable, address a recognised need and have verifiable impact.
Replacing the SPE Process. For larger pieces of work network engineering (e.g. 2+ months with multiple contributors), the onchain treasury process will still be used.
SPE Governance Structure
Custody: Funds held by the Livepeer Foundation.
Decision process: All decisions made live in weekly Review Team meetings via a simple 2-of-3 vote, with a written decision log published after each meeting. Any Review Team member with a direct financial interest in a decision must recuse.
| Role | Who | Responsibilities | Paid by SPE? |
|---|---|---|---|
| Review Team Lead & Technical Director | Rick Starr (Livepeer Foundation) | Refines scope of RFPs with Suggester; leads RFP team selection; signs off on technical quality for all RFP Track 1 work - sole gate on technical delivery, no broader Review Team vote required; produces pilot evaluation report | No |
| Reviewer, RFPs | TBD — Nominated by Orchestrators. Selected by Review Team Lead. | Reviews and scores RFP applications against evaluation criteria; represents community interests in first-pass scoping; casts an independent vote on Tranche 2 and Tranche 3 payouts; publishes written rationale for any dissent or recusal | Yes ($7,500 / 3 months, ~10 hrs/week) |
| Reviewer, Retro Grants | TBD — Nominated by Orchestrators. Selected by Review Team Lead. | Owns first-pass review of retroactive grant applications against the published rubric (supported by AI pre-screening); triages and quality-screens incoming applications; compiles monthly review shortlist; proactively surfaces strong unsubmitted community work; casts an independent vote on retroactive grant approvals | Yes ($7,500 / 3 months, ~10 hrs/week) |
| Roadmap Manager | Rich O’Grady (Livepeer Foundation) | Facilitates community Suggestion sessions; manages the Suggestions pipeline; promotes validated ideas to the roadmap; signals funding path per item | No |
| Program Manager | Mehrdad (Livepeer Foundation) | Runs and records weekly Review Team meetings; maintains the public decision log; tracks milestone status; manages remediation windows | No |
| Contributors | Community | Deliver scoped outputs with full ownership and accountability | Yes |
This SPE will have AI-native workflows embedded from day one, reducing Review Team overhead, keeping the public record current, and ensuring consistent quality control without manual lift.
AI-led tasks include: retro grant pre-screening; automating meeting notes and updates; decision log drafting; application triaging.
Timeline
15 May 2026 — Funding Programs Live
Review Team seated. Weekly cadence established. First 3 RFPs posted publicly. Retroactive grants open. Public tracker live. Key assessment: did we launch on time with the right scope?
30 Jun 2026 — First Deliveries Verified
At least 1 RFP delivered. At least 1 retroactive grant awarded. Key assessment: is the quality bar holding? Is the 14-day selection target being met?
30 Aug 2026 — Pilot Complete
Minimum 3 RFPs delivered, with aspirational goal of 5. All impact assessments done. All spend published. Evaluation report out with a clear recommendation: continue, modify, or stop.
Budget Breakdown
Total requested: $95,000 USD-equivalent (4-month pilot)
| Budget line | Assumption | Amount (USD-equiv.) |
|---|---|---|
| Review Team Member compensation (×2) | 2 × $7,500 / 3 months (~10 hrs/week) | $15,000 |
| Track 1 — RFP pool | Aim: 3 RFPs over pilot at ~$15k average | $45,000 |
| Track 2 — Retroactive Grants pool | Aim: 5 retroactive grants at ~$3k average | $15,000 |
| Funding Buffer — for additional grants or RFP | To either be allocated, returned to the onchain treasury or carried over to future SPE | $20,000 |
| Total | $95,000 (in LPT) |
If uptake for RFPs and/or retroactive grants are lower than expected at the 30 June checkpoint, the Review Team has the opportunity to reallocate between pools based on where community activity is strongest.
Any unspent funds or incomplete RFPs at pilot end are returned to treasury (with a small undefined, remediation window if work near completion).
All amounts sized using the 30-day moving average LPT price at the time of RFP approval.
Deliverables
Pilot success criteria (3 months):
1. Minimum 3 RFPs reach verified impact - At least 3 scoped engineering RFPs complete with a confirmed definition of done, verified by the Technical Director and community Review Team Members. Each delivery is published publicly with a written rationale and outcome assessment.
2. Minimum 5 retroactive grants funded - At least 5 retroactive grant applications reviewed and awarded, covering work that has already shipped and meets the published rubric. Each award is published publicly with written rationale, the amount approved, and the impact evidence submitted by the contributor.
3. Three public SPE updates published during pilot - One per month (June, July, August). Each update covers: RFPs active and delivered, retroactive grants reviewed and awarded, funds disbursed to date, and Review Team decisions with written rationale. Published within 7 days of the monthly Review Team meeting. Zero black-box decisions.
4. Pilot evaluation & impact report published by 30 Aug 2026 - A concise report covering: RFPs funded and delivered, retroactive grants awarded, total spend vs. budget, community signal on delivered work, and a clear recommendation: continue, modify, or stop. Written for a community audience, not an internal one.
Transparency and Accountability
- Every RFP and retroactive grant decision published publicly with written rationale.
- Public tracker RFPs: all RFPs, status, funds allocated and disbursed, Review Team decisions
- Public tracker grants: all retro grant applications, decisions, amounts awarded, and impact evidence
- Review Team Members publish justification for any overrule or recusal
- Monthly written reporting to the community (per SPE norms) plus update at Water Cooler
- All conflicts of interest disclosed and recused
Partial. One RFP issued, delivered and closed complete (
Seven decided, seven different builders, five of them bounties, each with a published rationale
Impact assessments are the grantees’ own published retros. The retroactive-grant retros are in, and the RFP’s impact is captured above. The two Direct Grants still in flight (Payment Clearinghouse, subgraph audit) and Gideon’s vllm-realtime bounty will publish theirs as the work completes.