Filecoin PGF roadmap
The filpgf.io roadmap
What we’re building into the Filecoin PGF portal, who is building it, and when each piece ships.
filpgf.io is becoming the place where the Filecoin Public Goods Funding (PGF) program shows its work: what’s funded, how those projects are doing, and how decisions get made. This page covers our objectives, the features that deliver them and the teams building them, how funding support comes together, and the roadmap for when each piece ships. For a shorter version, see the companion blog post.
What we’re building toward
The build is organized around three objectives: understand what funding produces, make applications and reviews faster and lighter, and give the people running the program one place to see and act on everything. Each one comes with a set of features, built together with our partners Karma, OSO, and Simocracy.
Objective 1
Measure impact and maintenance
Funding only matters if what it produces lasts. OSO tracks impact and maintenance for every funded project across Kernel, R&D, and Revenue Development, and is being built out to flag stalled data sources before anyone has to go looking. Verified outcomes will also be captured as project-level hypercerts.
- Metrics on cover pages and reports
- Automated health monitoring
- Hypercerts on verified outcomes
Objective 2
Make applications and reviews AI-native
Applications adapt to each project’s scope of work, and a Simocracy sim drafts the first review of every milestone. A person always makes the call, and every review is captured in filpgf.io.
- Adaptive applications
- Sim-assisted milestone review
- Human sign-off
Objective 3
Track every funded project, keep ops light
Karma gives the program team one admin view of every application, milestone, and invoice, queued by priority and age. Feedback from Slack, Telegram, and Simocracy attaches to the item it belongs to, so there’s less chasing for everyone.
- Priority and age queues
- Slack and Telegram feedback
- SLA reporting
Underneath all three: one source of truth for the program, so funded teams, reviewers, and the program team all work from the same information.
Who we’re building with
One loop, three partners
These pieces are being built to work together. Applications, reviews, and impact data all feed one system, so what we learn at each step informs the next. Every time the loop comes around, the next review starts with better data than the last one, and the same process helps us efficiently track and showcase the ecosystem’s needs and priorities, and keep applicants and stakeholders informed about them.
filpgf.io
One shared record
Syncs
applications & milestones
Informs
funded projects to track
Feeds back
impact & maintenance metrics
1 Hosts & surfaces
Karma
Holds the program’s records: incoming applications and milestones, plus the reviews and metrics the loop produces, surfaced in admin queues, cover pages, and Ask Karma.
Admin views →
2 Review
Simocracy
A sim drafts the first assessment of every application and milestone. A person reviews it and makes the call.
AI-native reviews →
3 Measure
OSO
Tracks impact and maintenance for every funded project, and flags a stalled data source before anyone goes looking.
Impact tracking →
How the pieces work together inside filpgf.io. Select a partner to jump to its feature area.
Feature spotlights
The features, up close
Each objective comes down to a set of features. Here is what each one does, what it looks like, and who is building it: OSO on impact and maintenance data, Simocracy on AI-native applications and milestone review, and Karma on the platform and admin views that tie everything together.
Funding a project is one decision. Knowing whether it keeps delivering takes ongoing measurement. In filpgf.io, every funded project gets two things kept current: impact (what the work is producing) and maintenance (whether it is still actively built and cared for). The approach runs across all three funding categories, with one addition for RevDev: a revenue-linked metric such as paid usage, onchain revenue, or paying customers.
What gets tracked, by funding category
| Kernel | R&D | RevDev | |
|---|---|---|---|
| Cover page with program-level and project-level metrics | |||
| Published health and maintenance metric per project | |||
| Automated monitoring for stale or failing data sources | |||
| Revenue-linked metric (paid usage, onchain revenue, or paying customers) | — | — |
Planned coverage by funding category. Goals: 90% of approved Kernel, RevDev and R&D projects carry at least one published health metric, and 100% of RevDev projects carry a revenue-linked metric. These are program goals, not yet results.
Kernel health monitoring, funded teams this batch
100%
922 of 922 expected reading periods carry a value
45
each on its own cadence, from a public source
14
Projects with a metric against them
| Team | Kernel functions by tier | Metrics | Functions | Coverage · 30d |
|---|---|---|---|---|
| Ankr | 1 Essential | 3 | 1 | 100.0% |
| Blockscout | 2 Essential | 6 | 2 | 100.0% |
| Chain.Love | 1 Essential | 5 | 1 | 100.0% |
| Curio / FilOz | 3 Essential | 5 | 3 | 100.0% |
| drand / Randamu | 1 Irreplaceable · 1 Essential | 5 | 2 | 100.0% |
| Filecoin Community Services / ChainSafe | 3 Essential | 3 | 3 | 100.0% |
| Filecoin Data Portal | 1 Essential | 1 | 1 | 100.0% |
| Filecoin Docs / FIL-B | 1 Essential | 2 | 1 | 100.0% |
| Forest / ChainSafe | 2 Essential | 2 | 2 | 100.0% |
| Goldsky | 1 Essential | 2 | 1 | 100.0% |
| IPNI / OIF | 1 Essential | 4 | 1 | 100.0% |
| Plumbline | 1 Essential | 1 | 1 | 100.0% |
| USDFC / Secured Finance | 1 Important | 4 | 1 | 100.0% |
| Zondax | 1 Essential | 2 | 1 | 100.0% |
Snapshot from the live Kernel monitoring view on filpgf.io, 25 September 2026. Every metric is fetched nightly from a public source by a pipeline the team does not control. Projects, the Kernel functions they steward, and their metric definitions are registered in the public pgf-monitor repository.
Automation keeps the numbers honest. Automated project health monitoring will flag when a source in the metrics dashboard goes down or drops below a threshold, so the team notices a stale number before the community does. Verified metrics then flow into cover pages, project views, bi-weekly and monthly reports, and the Ask Karma assistant. For Kernel, this already runs in public: the projects stewarding each Kernel function and their metric definitions are registered in the pgf-monitor repository on GitHub, which powers the Kernel health monitoring on filpgf.io.
Two processes go AI-native: how projects apply, and how they are reviewed. For both project applications and project milestones, a sim does the first assessment, and a person reviews it and decides if it’s cleared for approval. The point is to make the ops side of reviews lightweight, for reviewers and for the teams receiving funding. Simocracy is now live in our review process, and iterations are ongoing.
AI-native application process
Applications will be pre-populated using existing project information, so Kernel teams do not need to resubmit information already on file. Teams can review the pre-populated application, make any necessary updates, and submit.
AI-native milestone review
Sim reviewer assignment and assessment, milestone submission notifications to the human reviewer, reminder triggers, and escalation flags for reviews that exceed their turnaround time.
The two processes, step by step
-
Pre-filled application
From information on file
-
Team review
Update and submit
-
Sim assessment
Sims evaluate
-
Committee decision
Humans decide
-
Milestone submitted
Update filed in filpgf.io
-
Sim assessment
Assigned sim reviewer
-
Human reviewer
Notified, reminded, escalated
-
Decision recorded
Captured in filpgf.io
Feature area 3
Admin views and funding workflows
Built with Karma
Built in partnership with Karma, the admin view is the centerpiece for the team running the program. Reviews and payment items queue up by priority and age, so it is always clear what to pick up next. Feedback from reviewers and funded teams in selected Telegram and Slack channels, along with Simocracy feedback, is integrated automatically into those same work items. The aim is better turnaround times and a better experience for every stakeholder, with AI making the ops side of reviews lightweight.
How the pieces connect
-
Slack & Telegram
Reviewer & project feedback
-
OSO
Impact & maintenance data
-
Simocracy
Sim reviews & feedback
Platform partner for filpgf.io
-
Public views
Funding, projects, impact
-
Project & reviewer views
Status, reviews, invoices
-
Admin queues
Reviews & payments by priority
Simplified data flow. Illustrative, not an architecture diagram.
How it comes together
Funding support: faster decisions, fewer nudges
The three objectives above meet in the day-to-day experience of funding support. Sims prepare the first pass and admins work from one queue, so people spend their time on judgment calls rather than chasing updates. These are the service commitments filpgf.io is designed to support, helping us keep SLAs consistent across a decentralized environment:
- Project milestone reviewers are notified of their sim’s assessment within one business day.
- Standard milestone reviews are completed within 7 business days. Reviews that require changes or further discussion may take additional time, with a target of 14 business days.
- We’re targeting a two-week payout window for invoices on verified milestones.
Service commitments for funded projects
Reviewer notified
Standard milestone review
Review needing changes
Invoice payout
Project follow-ups
Overdue milestone updates get a nudge every 3 business days, then weekly after 10.
Reviewer follow-ups
Pending reviews follow the same cadence, so nothing stalls quietly.
Admin queues
Every application, milestone and invoice becomes a work item with a status and an age.
The service model filpgf.io is designed to support. Admin views will report SLA performance every two weeks.
Feedback from Simocracy and from selected Slack and Telegram automations will be incorporated through successive iterations of this build-out.
When each feature ships
Here is what we plan to ship and roughly when, in three delivery windows. The roadmap is rolling: we revisit priorities every week, so sequencing and dates may shift as we learn from funded teams, reviewers, and admins.
-
In progress
Funding framework, transparency & AI-native operations
-
Funding metrics everywhere
Metrics land in cover pages, individual project views, and bi-weekly/monthly reporting, and are available through Ask Karma.
Metrics & impact
-
Simocracy integration v1
Selection Committee members’ sims are up and running.
Sims & reviews
-
Automated milestone review
Sim reviewer assignment and assessment, notifications to the human reviewer, reminder triggers, and escalation flags for slow reviews.
Sims & reviews
-
AI-native application process v1
Application intake, AI-assisted application development and review, and a submission workflow.
Applications
-
Admin queues live
Applications, milestones and invoices each get a work item with trackable status and age.
Admin operations
-
Navigation for the new funding framework
UX and navigation improvements that support the three funding categories.
UX & comms
-
-
Next
Funding ops data & automation
-
Keep iterating
Continued iterations of Simocracy, application intake, admin queues, and metrics engagement (Ask Karma, reporting, cover pages).
Sims Apps Admin Metrics
-
Slack & Telegram integration v1
Relevant project discussions connect to the matching admin work items and project records.
Integrations Admin operations
-
Automated project health monitoring
Flags when a source in the metrics dashboard is down or below threshold.
Metrics & impact
-
AI analytics dashboards for admins
Admins can sort and view how AI assessments of milestones and applications line up with human reviewer sign-off.
Admin operations Sims & reviews
-
-
Later
Integrated program intelligence
-
Keep iterating
Continued iterations of Simocracy, application intake, admin queues, and metrics engagement.
Sims Apps Admin Metrics
-
Project-level Hypercerts
Project-specific hypercerts that capture and showcase verified outcomes and impact from funded work.
Metrics & impact Integrations
-
Showcase roadmap impact
Blogs and content highlighting what shipped through the filpgf.io roadmap and the impact it delivered.
UX & comms
-
What this adds up to
If the next six months go to plan, a community member can see where PGF money went and how those projects are doing without asking anyone. A funded team can check a milestone and an invoice without sending an email. A reviewer can open one page, see a sim’s draft beside the full application, and decide. And the team running the program can spot a stalled project or a failing data source early, and report on it without assembling a spreadsheet first.
We will keep sharing progress as each window closes. In the meantime, explore the live portal and tell us what would make it more useful for you.
Roadmap windows are indicative and may change as priorities evolve. For the short version of this page, see the filpgf.io blog post.