Skip to content

Sprint - June 1 to June 12 #60123

New issue
New issue
Closed
Closed
Sprint - June 1 to June 12#60123
Labels
sprintSprint planning

Description

@ioannisj
ioannisj
opened on May 26, 2026

Sprint Planning

List of incidents + any action items that haven't been scheduled yet

Team sprint planning

For your team sprint planning copy this template into a comment below for each team.

# Team ___

**Support hero:** ___
**Off during the sprint:** ___

## Quarter goals

1. Quarter goals, status (red/yellow/green) and action points if yellow/red

## Retro

### Tasks

<!-- Grab the high and low priority items from last time and add whether that item was completed or not -->

## Plan

### High priority

-

### Low priority / side quests

-

### Are any other teams impacted by this plan? If so, tag them here

-

Activity

  1. ioannisj
    added
    sprintSprint planning
    on May 26, 2026
  2. ioannisj
    pinned this issue on May 26, 2026

  3. ioannisj commented on May 26, 2026

    @ioannisj
    ioannisj
    on May 26, 2026
    ContributorAuthor

    Team Client Libraries

    Support hero: @turnipdabeets then @dustinbyrne
    OOO: 1 day @ioannisj , 1 day @marandaneto (and + 1 day attending a conf.)

    Quarter goals

    • 🟡 Goal 1: What makes PostHog more tailored to how you actually use it
      • ⚪ Discover SDK + default gaps
      • 🟡 Proactive user research (interviews, surveys, mobile/backend focus)
      • 🟡 Promoted properties
    • 🟡 Goal 2: Agent-friendly SDKs
      • 🟡 Documenting all SDKs with jsdoc style
      • 🟡 Adding all SDKs docs to context-mill
      • 🟡 Updating all SDKs on posthog.com/docs
      • 🟡 Removing and/or promoting outdated beta/exp/alpha features on all SDKs
    • 🟡 Goal 2: PostHog libraries vNext
      • ⚪ Offline-first / async-first posthog-js
      • 🟡 Distributed tracing headers + remote config support
        • 🟢 Client tracing headers sent on mobile
        • 🟢 Server-side SDKs supporting request contexts read tracing headers
        • 🟢 All server-side SDKs support request contexts where applicible
        • 🟡 All mobile and web clients support first class tracing headers
      • ⚪ More framework-specific wrappers
      • 🟡 Mobile session replay (triggers, min duration, RN/Flutter/macOS, perf)
      • 🟡 Mobile error tracking (iOS improvements, RN support & restructuring, breadcrumbs)
      • ⚪ Flutter desktop support (Windows/Linux)
    • 🟡 Goal 3: Maintaining the SDKs
      • 🟡 SDK test harness + CI integration
      • 🟢 SDK rotation ownership
      • 🟢 Config standardization (API key/token)
      • ⚪ Ingestion rate limiting support
      • 🟢 Unified remote config endpoint (/static/token/config)
      • 🟢 SDK design guidelines and best practices
    • 🟡 Goal 4: Log for Mobile
      • 🟢 Add support for collecting and sending app-side logs from Android
      • 🟢 Add support for collecting and sending app-side logs from iOS
      • ⚪ Add support for collecting and sending app-side logs from Flutter
      • 🟢 Add support for collecting and sending app-side logs from React Native

    Retro

    Details

    High priority

    Low priority / side quests

    Plan

    High priority

    • Android surveys default UI delegate @ioannisj
    • Mobile error tracking exception steps (iOS, Android, RN, Flutter) @ioannisj
    • Session replay event triggers for Flutter @ioannisj
    • manual capture API for Logs - Flutter @turnipdabeets
    • first class tracing header support in posthog-js (i.e., no internal flags) @dustinbyrne
    • initial stubs of offline-first / async-first posthog-js @dustinbyrne
    • Cont. Make SDKs agent-friendly @marandaneto
      • document all public APIs and configs (source code)
      • make sure that all public APIs and configs are documented (posthog.com / docs)
      • make sure that context-mill has access to the above
      • promote or remove all beta/alpha/exp features

    Low priority / side quests

    🟢 =finished 🟡=in progress 🔴=won't finish ⚪=not started

  4. tomasfarias commented on May 27, 2026

    @tomasfarias
    tomasfarias
    on May 27, 2026
    Contributor

    Team Batch Exports

    Support hero: @tomasfarias / @rossgray

    Quarter goals

    • Objective 1: Unify batch exports and real-time destinations
    • Objective 2: Reduce pipeline architecture technical debt
    • Objective 3: Self-serve debugging

    Retro

    Details

    ⚪ CDP backfilling

    • Rollout as beta
    • Add usage tracking so we're aware when users start using it
    • Add docs to CDP pages

    🟡 Split up S3 destination into supported destinations into:

    • S3-compatible (catch-all)
    • AWS S3
    • Cloudflare R2
    • GCS
      ⚪ Support new auth methods for S3 and GCS via Integration model

    🟡 Integration for PostgreSQL

    • migration of existing config

    🟡 File download batch export feature

    • Deployed, but encountered some permission issues
    • Added a SKILL + MCP tool support

    ✅=finished 🟡=in progress 🔴=won't finish ⚪=not started

    High priority

    • CDP backfilling

      • Rollout as beta
      • Add usage tracking so we're aware when users start using it
      • Add docs to CDP pages

    • Split up S3 destination into supported destinations into:

      • S3-compatible (catch-all)
      • AWS S3
      • Cloudflare R2
      • GCS

    • Support new auth methods for S3 and GCS via Integration model

    • Integration for PostgreSQL

      • migration of existing config

    • File download batch export feature

      • Fixing a couple last minute bugs, but ready to rollout very soon.

    Low priority

    • Resource based allocator + auto-scaler for batch export workers

  5. langesven commented on May 27, 2026

    @langesven
    langesven
    on May 27, 2026
    Contributor

    Team Cloud Platform

    Quarter goals

    1. Quarter goals, status (red/yellow/green) and action points if yellow/red

    🚀 Make deployments great - Driver: Sven Lange
    Increase transparency, speed, and maintenance of our deployment pipeline. Everyone knows what versions are deployed where. Easy!

    Retro

    Details
    • merge SREHog into SherlockHog @k33lan
    • infra-dev base setup @Z3r0Sum
      • Fixes to internal Argo CD along the way:
        • Duplicate public Route 53 zone -> cleanup up
        • Proper management of Cognito Secret - was manually created via lens
        • gRPC based ALB that properly handles CLI requests from our deployment pipelines / locally if desired -> was causing sporadic 464 failures for deployments.
    • Move our deployment pipelines on the charts repo towards depot - progress testing with the shadow pipelines @Z3r0Sum
    • Wrap up posthog-<admin|query|report>-django migration to posthog-app chart @langesven
    • Stand up posthog-toolbox-django with posthog-app chart, still parallel to old one for validation, made it the default toolbox to use @langesven
    • Establish prune: true as the default setting for all ArgoCD app-of-apps to auto-remove deleted resources @langesven
    • Fix perpetual bug with long commit messages breaking the charts repos state.yaml / subsequent deployment flows @langesven
    • Upgrade Karpenter to v1.9.0 in dev/prod-eu/prod-us, aligned with managed-warehouse now @langesven

    Tasks

    Plan

    High priority

    • Finish migrating posthog-django stack to posthog-app chart @langesven
    • Migrate remaining apps to posthog-app chart @langesven
    • Productize SherlockHog @k33lan
    • ArgoCD upgrade, EKS upgrade @Z3r0Sum
    • More actionable migration notifications/failures @Z3r0Sum

    Low priority / side quests

    • decomission old Victoria Metrics cluster @k33lan
    • make Karpenter actually consolidate nodes / improve utilization of Kubernetes nodes @langesven

  6. mariusandra commented on May 27, 2026

    @mariusandra
    mariusandra
    on May 27, 2026
    Collaborator

    Team Data Tools

    Q2 2026 objectives

    Objective 0 – Make the MCP a first-class, competitive data interface (George, Marius)

    • 🟢 End-to-end querying workflows: Create, run, iterate on, and debug SQL/HogQL queries with strong error messages, metadata, and performance guidance.
    • 🟡 Data discovery & context: Explore connections, schemas, properties, and other metadata. Provide guidance on which connections/tables/etc to use.
      • todo: properties
    • 🟢 Cross-database querying: Seamless querying with HogQL across ClickHouse, Postgres, DuckDB, and other sources with consistent ergonomics
    • 🟡 Visualization & reporting: Generate charts, choose appropriate visualizations, and create reusable outputs like dashboards and reports
      • todo: use the right viz
    • ⚪ Notebook & workflow integration: Full notebook support via MCP, including creation, editing, and async/background workflows
    • ⚪ Analysis & intelligence: Using notebooks, answer business questions, run statistical analyses, detect inconsistencies, and optimize queries beyond syntax

    Objective 1 - Build a BI experience that feels competitive and joyful (Marius)

    • 🟡 New chart types, prioritizing the most-requested visualizations
    • 🟡 SQL editor reliability and UX improvements
    • ⚪ Better variables, including multiselect, query-backed and time-based variables (e.g. for dashboards)
    • ⚪ Easier schema exploration and an alternate “no-sql” query building interface (manual controls + AI integration).
    • ⚪ Ability to click in and out of rows to explore deeper.

    Objective 2 - Make notebooks the default way to analyze data (Anna)

    • ⚪ Python notebooks for customers
    • ⚪ Core notebook UX and stability improvements
    • ⚪ Better integration between notebooks, HogQL, DuckDB, and external databases
    • 🟡 Initial collaboration improvements for notebooks (lock nodes if someone is editing, etc)
    • ⚪ A clearer default analysis flow centered on notebooks

    Objective 3 - Build a first-class querying and DuckDB experience (George)

    • 🟢 First-class DuckDB support (not treated as “just Postgres”)
    • 🟢 Better query errors so failures are understandable and actionable
    • ⚪ Improvements to the HogQL type system to support earlier errors
    • ⚪ Calculated properties and lambdas on top of HogQL tables
    • ⚪ Improved UX for connecting to databases

    Retro

    Plan

    Support rota: George then Marius

    High priority

    • MCP reliability. Even better MCP querying @georgemunyoro
    • Project MNSL - Make Notebooks Suck Less
      • AI first notebooks

    Low priority / side quests

    Are any other teams impacted by this plan? If so, tag them here

    • N/A

  7. rafaeelaudibert commented on May 27, 2026

    @rafaeelaudibert
    rafaeelaudibert
    on May 27, 2026
    Member

    Team Growth

    Support hero: @rafaeelaudibert and @MattBro
    Off during the sprint:

    • @rafaeelaudibert will be in SF from 3rd-9th, then Sao Paulo from 10th-11th, off 12th

    Quarter goals

    1. Make agentic sign-ups first-citizen of PostHog 🟡 (@MattBro)
    2. Allow anyone to integrate/provision an org + project in PostHog 🟡 (@rafaeelaudibert )
    3. Activation rate across all onboarding experiences should be 60%+ 🟡 (@fercgomes)
    4. Shape the external interface for Signals 🔵
    Details ⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned

    Plan

    High priority

    • Fix the "new product" ads in the sidebar, make it look much nicer (endpoints and something something warehouse are launching next week) @rafaeelaudibert
    • Marketplace: Implement OAuth RFC + document how the whole flow works + plan on marketing it @rafaeelaudibert @MattBro
    • Improve cross-selling, do not use only the sidebar, abstract away from the "UserProductList" @rafaeelaudibert
    • Wizard + onboarding sync @fercgomes
    • Standardize empty state for all products @fercgomes
    • Wizard provisioning in the website landing page POC @MattBro

    Low priority / side quests

    Retro

    Details

    Low priority / side quests

    What else did we do?

    • A bunch of small PRs to PostHog Code while @fercgomes is trying to become a power user, there's 95 of them (probably)
    • AIO + Error tracking empty state (@fercgomes )

  8. jonmcwest commented on May 27, 2026

    @jonmcwest
    jonmcwest
    on May 27, 2026
    Contributor

    Team Logs

    Support hero: @DanielVisca and @jonmcwest
    Off during the sprint: @frankh is back on June 8th

    Quarter goals (ends June 30th)

    Product autonomy

    • 🟢 Can use logs without the UI (e.g. via Max)
    • 🟡 Detect and send interesting signals, e.g. anomaly detection, spikes, errors
    • 🟡 APM
    • 🟡 Traces
    • 🟡 Metrics
    • 🟡 Reduce migration blocker feedback to 0

    Table stakes

    • 🟡 Solid backups @frankh
    • ⚪ Quota limits so complex queries don't impact other users (e.g. searching for "1" in team2)
    • 🟡 Retention
    • 🟢 Alerts
    • 🟢 Saved views / dashboards
    • 🟢 Context view
    • ⚪ Archival storage
    Details ⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned

    Retro

    Details

    High priority

    • 🟢 traces out for posthog internal use in prod @frankh
    • 🟢 5000 alerts running with no performance issues @jonmcwest
    • 🟢 service level rate limiting @DanielVisca
    • 🟢 simple ingestion controls e.g. log level, or sampling, or both @DanielVisca
    • 🟡 Alerts hooked up to signals @jonmcwest
    • 🟢 improve traces MCP (feature flag it) @DanielVisca
    • 🟢 cut over clickhouse clusters to ec2 ones @frankh
    • 🟢 sort out ingestion autoscaling in prod-us @frankh

    Low priority / side quests

    • send billing metrics for retention

    Plan

    High priority

    Low priority / side quests

    Details ⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned

  9. rorylshanks commented on May 27, 2026

    @rorylshanks
    rorylshanks
    on May 27, 2026
    Contributor

    Team ClickHouse

    AKA Team Read Fewer Bytes Please

    Retro

    Details

    High priority

    Low priority / side quests]

    🟢 =finished 🟡=in progress 🔴=won't finish ⚪=not started

  10. thmsobrmlr commented on May 27, 2026

    @thmsobrmlr
    thmsobrmlr
    on May 27, 2026
    Collaborator

    Team Product Analytics / MCP Analytics

    Topics

    • MCP analytics / @lucasheriques joining:
      • Next steps for MCP analytics.
        • → Move SDK code into posthog-js; capture session intent for team 2; evaluate product from there + customer calls. Probs worth discussing how much of observability vs. analytics.
      • @lucasheriques' ongoing responsibilities for surveys.
        • → Maintenance mode; maybe move to conversations product?; re-evaluate in 2-3 months.
      • Whether @lucasheriques will focus only on MCP analytics, or also help with regular product analytics work, for example the support queue.
        • → Keen to do so; Focus on MCP analytics first.
      • If he will help with broader product analytics work, some onboarding would probably make sense.
        • → Offsite/onboarding later this year? Maybe together with analytics platform and query performance.

    Past sprint

    Details * @thmsobrmlr * MCP improvements: Exposed system.dashboard_tile via HogQL; added validations for the test account filters setting; added MCP tool callouts to docs * Started working on "compare against previous" for funnels * Customizable colors for SQL insight results * A bit of migration to the product analytics folder.. * @sampennington * Week of Support * fixed bug with enable_analyzer + custom hogql filters failing * fixed mcp creating invalid insights in notebooks * Fixed a bunch of query errors identified from new metabase dashboard. Now tagging queries better to improve alerting * Finished migration and released hog-charts for trends and stickiness insights. Funnels, Lifecycle, Retention WIP, probably done by eow will update! * Created BoxPlot, Pie Chart, Sparkline chart components * @gesh * Support week 1 * Fixed stacked bars tooltip showing per-series data, instead of cumulative * Retention charts can now be configured to start from D1, instead of D0 * Fixed LineGraph x-axis to display properly custom labels * Added tooltip datetime header in funnels (historical graph) * Prepared HogQL to aggregate both mobile and web events into a chart * Display long URLs in HogQL tables on a few lines * Investigated possible issue with tracking converted/dropped-off users in funnels * A few more tickets about identity configs, trends drop * Some devex /hogli stuff around properly loading of env variables * [MCP Analytics] Sessions view is live * Sessions listing released * Added session intent generation * @mjwarren3 * Analysis into events breakdown MoM (Autocapture vs. Pageviews vs. Custom events, etc.) - To include in next growth review * First PR submitted (show annotations toggle) * Demo for data warehouse-sourced events

    Plan

    • @mjwarren3
      • Growth review prep - We're doing with exec team next month
      • Proposal for PostHog Plus plan
      • Run another batch of customer interviews
      • Plan next team offsite
      • DWH-sourced events (if time)
    • @sampennington
      • Move hog-charts into Quill
      • Migrate MCP Apps to hog-charts
      • Week of support
    • @gesh
      • Support week 1

  11. robbie-c commented on May 27, 2026

    @robbie-c
    robbie-c
    on May 27, 2026
    Member

    Team Query Performance

    Support hero: @andyzzhao
    Off during the sprint: @aspicer

    Quarter goals

    Query performance across the app needs to improve to support our largest customers.

    Finish backfilling sessions v3 on the new cluster
Fix the gaps in our query tagging
Ensure that new teams are created with an internal/test user cohort and help big customers migrate onto this
Explore person ID sharding and help other teams use preaggregation
Improve custom query performance with DMAT columns / and EAV table [Sandy Spicer](https://posthog.com/community/profiles/30369)
Improve the query codebase [Sandy Spicer](https://posthog.com/community/profiles/30369)

    Event schema Sandy Spicer

    Continue improving this feature

    Add support for blocking events that don't match the schema

    Event schema Andy Zhao

    The taxonomy filter contributes to posthog feeling slow, and we can also help out posthog-ai by keeping the taxonomy recent, accurate, and fast

    Expand properties-rs to capture property values as well as property names
Ensure that posthog-ai has access to a good quality taxonomy

    Retro

    Tasks

    High priority

    @andyzzhao

    • continue on ingestion side of new properties table 🟠 - write amplification is too high, clickhouse can't keep up, scale up cluster or dedup during ingestion or filter out high cardinality properties
    • testing ^ (maybe do a backfill)

    @mjwarren3

    • may query performance review - should be this coming sprint

    @robbie-c

    • merging the new rust parser and deploying in shadow mode 🟢
    • block untagged queries for team 2 🟠

    @aspicer

    • dmat stack merged 🟡 - some got merged, unsure if whole stack did
    • monitor offset removal

    Low priority / side quests

    @robbie-c

    • (stretch goal) run local autoresearch and try to ship some PRs 🔴
    • finish blog posts 🔴

    Plan

    High priority

    @andyzzhao

    • continue on ingestion side of new properties table
    • testing ^ (maybe do a backfill)
    • query log inside posthog

    @mjwarren3

    • may query performance review

    @robbie-c

    • monitor the new rust parser in shadow mode

    Low priority / side quests

    @robbie-c

    • (stretch goal) run local autoresearch and try to ship some PRs

  12. meikelmosby commented on May 28, 2026

    @meikelmosby
    meikelmosby
    on May 28, 2026
    Collaborator

    Team Workflows

    Time Off

    Support Hero

    Retro

    High priority

    • new consumer for CDP & stability work / scaling stuff @benjackwhite

    • wrap up incident @meikelmosby

      • ship replay capabilities for blocked runs @meikelmosby
      • publish public post mortem
      • follow up comms with impacted customers

    • alert customers when they get rate limited (PR) @meikelmosby

    • Move us over to only use job-queue v2 for workflows @meikelmosby

    • Ship Email delivery throttle (depending on job-queue-v2 work) @meikelmosby

      • dedicated email queue PR
      • ses rate limiter PR
      • email send retries PR

    • "Wait until event" workflow action / conversion goal for team2 @dmarchuk

    Low priority

    • emit email events and make them opt-in @meikelmosby
    • LLM workflow node @dmarchuk
    • Merge push only for team2

    This Sprint

    High priority

    @dmarchuk

    • Wait until

      • first ship it for team2 for testing
      • roll out across every team

    • emit email events as posthog events to react upon

      • signing of emails
      • logging of email-events like rejected, bounced etc (failure cases)

    • Push Notifs

      • team2 rollout and heavy testing
      • making it GA

    --

    @meikelmosby

    • exclude http status codes from emitting failure metrics

    • high vuln security fixes

    • migration to new google ads api bc of deprecation

    • Ship Email delivery throttle

      • dedicated email queue PR
      • ses rate limiter & email send retries PR
      • Dont' let heavy use teams dominate the queue

    • Behavioral cohorts on team 2 (internal testing)

    • support encrypted secrets for workflow inputs

    --
    @mayteio

    • Workflow MCP (CRUD & debug)

      • table stakes (CRUD, archive not delete, only when in draft, testing, etc) for workflow CRUD
      • debugging of workflows using their metrics, logs, etc
      • Batch workflow blast radius & run/scheduling
      • email creation
      • read-only workflow builder UI in the mcp

    • Destination MCP (CRUD & Debug)

      • filters, inputs, input_schema, hog code generation, testing, bla bla
      • debugging (querying metrics, logs, etc)

    • Implement revisions for workflows (V1)

      • creating revisions
      • backfilling (one off migration)
      • using versioned workflows in the executor

    Low priority

    • cleanup grafana from dead consumers @meikelmosby
    • alert customers when they get rate limited (PR) @meikelmosby
    • move hogMasking to PG @meikelmosby
    • move existing redis usage to valkey @dmarchuk
    • RFC for LLM Nodes / Agent use @dmarchuk
    • re-run of invocations (take over Bens work)
    • opt out list CRUD api being able to be used via api-key
      • download/upload via csv
      • mcp support

  13. vdekrijger commented on May 29, 2026

    @vdekrijger
    vdekrijger
    on May 29, 2026
    Contributor

    Team Analytics Platform

    Support hero: @MattPua
    Off during the sprint:
    N/A

    Legend

    ⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned 🟣 = Ongoing

    Quarter goals
    Dashboard stability and enhancements @MattPua
    • 🟡 New and expanded tests to prevent unexpected failures
    • 🟡 Collection of UI/UX changes that make working with dashboards a smoother experience
    • 🟡 Better controls and components beyond showing just graphs
    Improve platform stability and reliability @vdekrijger
    • 🟢 A clear definition of "success" for alert checks, subscriptions, exports, and query service
    • 🟡 Stability improvements to better deal with degraded infrastructure
    • 🟡 A 99.95% success rate for our background jobs over the last 28 days rolling
    • 🟡 Anomaly detection to proactively spot regressions
    AI-first support for our owned components @vdekrijger
    • 🟡 Signal integration for specific alerts
    • 🟡 AI suggestions during creation
    • 🟡 Improved AI onboarding
    • 🟡 Investigate ways to deal with adjusted usage patterns from AI

    Retro

    1. Loved:
      What went well during the project? These are the things that team members appreciated, enjoyed, or found to be effective.

    2. Loathed:
      What made things worse? What was missing or needs improvement? This category identifies gaps, such as missing resources, skills, or support that would have made the sprint more successful.

    3. Longed for:
      What do team members wish they’d had during the sprint? This could be anything that would have made the work more efficient or enjoyable, like better tools, clearer requirements, or more time.

    4. Learned:
      What new knowledge or skills did the team gain? This could be technical knowledge, better communication methods, or insights into the project or process.

    Retro notes here: https://posthog.slack.com/docs/TSS5W8YQZ/F0B741LSUVA?focus_section_id=temp:C:Ace8f4220fa108e4789928fc6056

    Tasks

    @MattPua

    • 🟢 High Frequency alerts (15 min intervals)
    • 🟡 Dashboard widgets

    @vdekrijger

    • 🟡 Prompt based subscriptions
    • 🟡 Proactive insight / pulse page to catch up on what happened
    • 🟢 Moving (a limited amount) of subscriptions to the free plan
    • 🟢 A bunch of terraform updates / PR's from OSS contributors.
    • 🟢 Fully deprecated the old alert flow in Celery

    Plan

    High priority

    @MattPua

    • finish off Dashboard widgets
    • Tackle some off the signals our team has generated
    • Support Hero rotation

    @vdekrijger

    • Want to finish off prompt based subscriptions
    • Make a lot of progress with the proactive insights
    • If there is time left, look into MCP changes for dashboard-create

    Low priority / side quests

    Are any other teams impacted by this plan? If so, tag them here

  14. hpouillot commented on May 29, 2026

    @hpouillot
    hpouillot
    on May 29, 2026
    Contributor

    Team Error Tracking

    Retro 🟢 🔴 🟡 🚫

    @hpouillot - Code owners assignement 🟡 Cymbal work
    @hpouillot - Distributed symbol set 🟢 Test in production
    @hpouillot - Embedding analysis (distance / monitoring / pipeline design) 🚫

    @cat-ph - Deep linking issue fixing 🟢
    @cat-ph - MCP Eval 🟢
    @cat-ph - Side quest: Rust SDK 🟡

    @ablaszkiewicz - sourcemap recommendation. Sync with wizard 🟡
    @ablaszkiewicz - project rate limiting deployed to everyone. 🟡 need approval
    @ablaszkiewicz - cleanup FF for insights 🟢

    @fivestarspicy - documentation & artifacts (how do we integrate with agent stack) 🟢

    OKR

    Recommendations - Setup & Priorities
    Grouping - Grouping
    Agent experience - MCP & Wizard
    Pricing exploration - Value alignment
    Issue Report - Signal integration
    Search query - Migration & Performance

    High priority

    @ablaszkiewicz - wizard analytics and follow-up
    @ablaszkiewicz - wizard android & flutter & ios
    @ablaszkiewicz - automatic recommendations generation & signal generation
    @ablaszkiewicz - rate limiting monitoring

    @cat-ph - Rust SDK sourcemap less
    @cat-ph - + sourcemap
    @cat-ph - MCP support for Alerting setup

    @hpouillot - Embedding analysis (distance / monitoring / pipeline design)
    @hpouillot - Distributed symbol set follow-up
    @hpouillot - Embedding Grouping

  15. Radu-Raicea commented on Jun 1, 2026

    @Radu-Raicea
    Radu-Raicea
    on Jun 1, 2026
    Member

    Team AI Observability

    Quarter goals

    Q2 goals are changed because of various changes in priorities/team org.

    Concrete goals we would like to tackle next (we didn't put the already done stuff):

    New quarter goals

    Retro

    Details

    A lot of the retro doesn't make sense because this is our first sprint planning in the new formation.

    Plan (W23, W24)

  16. haacked commented on Jun 1, 2026

    @haacked
    haacked
    on Jun 1, 2026
    Contributor

    Team Feature Flags

    Support hero:

    Off during the sprint: @gustavohstrassburger out Tue 06/09 and Wed 06/10

    Quarter goals

    Link to goals

    1. Isolated flags-specific infra 🟡
    2. Load testing framework ⚪
    3. Sub-100ms P99, consistent response 🟡
    4. Decouple flag evaluation from persons DB 🟡
    5. Delete all /decide code 🟢
    6. Split remote_config from /flags 🟢
    7. Get /local_evaluation in a good state 🟡
    Details ⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned

    Retro

    The two flags teams became one this sprint: Flags Platform merged into Feature Flags across the codebase, the handbook, and the project board automation. On the infra side, the billing aggregator finished its move to authoritative mode in both prod-US and prod-EU, backed by accurate metrics and atomic counters and gated by a cutover switch. SDK consistency was a big theme: a single $feature_flag_called dedupe key now spans eight SDKs, and React Native property types are preserved through flag evaluation. We shipped bot blocking end to end, kept pushing on cache instrumentation and Redis config warming, continued the legacy super_groups teardown (both read and write paths, plus the frontend UI), and dual-wrote the remote config hypercache to shared Redis.

    Details

    @haacked

    Billing aggregator → authoritative:

    Flag evaluation & cache:

    Observability & audit:

    Team merge & board automation:

    SDK:

    @gustavohstrassburger

    $feature_flag_called dedupe key (cross-SDK):

    Flag dependency conditions:

    Cohort calculations:

    Other:

    @matheus-vb

    Bot blocking:

    Cache instrumentation & config:

    Cohorts:

    @patricio-posthog

    Legacy super_groups removal:

    Remote config & key lifecycle:

    Perf & alerts:

    Plan

    Project Board

    High priority

    @haacked

    @gustavohstrassburger

    @matheus-vb

    @patricio-posthog

    Side quests

    • None this cycle.

  17. Daesgar commented on Jun 3, 2026

    @Daesgar
    Daesgar
    on Jun 3, 2026
    Contributor

    Team Cloud Foundations

    Support hero:

    Retro

    High Prio

    • 🟡 tailscale setup - final rollout + monitoring and alerting @MichaelKutsch-ph
    • ✅ smokescreen - review alerting @Daesgar
    • 🟡 rework security group rules + golden helm chart integration for RDS access @Daesgar
      • ✅ dev finished and working
    • ✅ Cilium: finalize EU rollout @danielxnj
      • only 3 PRs pending to be merged today
    • ✅ SOC2 point in time evidence work @danielxnj
    • ✅ tagging strategy @sinthetix
    • ⚪ TF drift detection @sinthetix

    side quest

    ✅=finished 🟡=in progress 🔴=won't finish ⚪=not started

    OKR Infra

    High Prio

    • 🟡 tagging strategy implementation @sinthetix
    • WarpStream renegotiation @sinthetix
    • 🟡 rework security group rules + golden helm chart integration for RDS access @Daesgar
      • prod-eu and prod-us
    • rework security group rules + golden helm chart integration for ClickHouse access @Daesgar
    • Aurora Postgres / RDS Terraform code consolidation @Daesgar
      • Upgrade to latest version of the module
      • Only create SG inside the module (don't allow passing other SGs to attach to the DB)
      • Bind rds-proxy creation to this module as well
    • Cilium: finalize US rollout @danielxnj
    • SOC2 population and samples work @danielxnj
    • 🟡 tailscale setup - final rollout + monitoring and alerting @MichaelKutsch-ph

    side quest

  18. lricoy commented on Jun 8, 2026

    @lricoy
    lricoy
    on Jun 8, 2026
    Member

    Team Web Analytics

    Legend
    ⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned 🟣 = Ongoing

    Support hero:

    @jordanm-posthog

    Retro:

    Details

    🟢 @arthurdedeus user interviews with AE's and TAM's for the CSP pivot
    🟢 @lricoy write the docs for bot analytics current usage and ask people to try it out
    🟢 @lricoy add real time and bot analytics tooling to the MCP
    🟢 @lricoy figure out the increased CPU usage on sessions query and use Robbie's precomputation
    🟢 @lricoy move the traffic-sim to open source posthog repo
    🟢 @jabahamondes drill down ad groups and ads and other utms that can be valuable
    🟢 @jabahamondes @lricoy try using the precomputation / reuse intermediate results
    🟢 @jabahamondes add Marketing Analytics tooling to the MCP
    🟢 @jabahamondes GSC integration
    🟢 @jordanm-posthog auto resolution for health issues on the wizard
    🟢 @jordanm-posthog give some love to the heatmaps feedbacks

    Low priority

    High priority

    @jordanm-posthog add achievements to the web analytics onboarding
    @jabahamondes MCP marketing analytics tuning
    @jabahamondes figure out roadmap for MA direction
    @arthurdedeus ai mode for customer analytics
    @arthurdedeus figure out account creation automations
    @arthurdedeus annoy everyone to use the product to get feedback
    @lricoy rollout precomputation for everyone
    @lricoy launch with Natalia and Ian the bot analytics details
    @lricoy do user interview on bot detection/analytics
    @mjwarren3 figure out eval metrics for web analytics mcp
    @mjwarren3 @lricoy user interview on web analytics

    Low priority

    @jordanm-posthog map heatmaps flows, failures and what we can do to make it good
    @lricoy release the other mcp tools
    @lricoy run an exploration on mcp-only web analytics

  19. pl commented on Jun 8, 2026

    @pl
    pl
    on Jun 8, 2026
    Contributor

    Team Ingestion

    Retro

    Details

    High priority

    • Isolate ingestion kafka topics
      • 🟡 Clean up ingestion producers @pl
      • 🟢 Add blackhole topic in EU @eli-r-ph
    • PersonHog
    • Expand first point of capture to support better processing tooling
    • Set ingestion SLAs and metrics and commit to them
      • 🔴 Break out the heatmaps pipeline @pl
      • 🟢 Break out the client warnings pipeline code @pl
      • 🔴 Set up latency metrics for all pipelines @pl
      • 🔴 Experiment with the SLOs dashboard @pl
      • 🔴 Change $ai_* events to use read-only person processing @pl
    • Improve per-partition ingestion throughput
    • Deduplicator
      • 🔴 Catch up with CH on their deduplication implementation
    • WarpStream
    • Infrastructure cleanup
      • 🟢 Move errortracking to its own namespace @z0br0wn
    • Postgres
      • 🟢 Improve postgres error handling (especially on pgbouncer scaledowns) @pl
      • 🟢 Review pgbouncer autoscaling @pl

    Incident follow-up

    • Look into deployments being stuck due to PVC issues

    Support follow-up

    • 🟢 Fix a bug in heatmap routing in capture @z0br0wn
    • 🟢 Batch import worker fixes @z0br0wn

    Side quests

    • 🟢 Switch over error tracking to PersonHog @z0br0wn
    • 🟢 Implement v1 endpoints for capture-internal @eli-r-ph
    • 🟢 Fix the harness for capture v0 @eli-r-ph
    • 🟡 Make a v1-compatible Rust SDK @eli-r-ph
    • 🟢 Property defs fixes / optimizations @eli-r-ph
    • 🟡 New superday @pl
    • 🟢 Selective CI for Rust @nickbest-ph
    • 🟢 Compress some Warpstream topics @jose-sequeira
    • 🟢 PersonHog optimizations @PostHog/team-ingestion

    OKRs

    • 🟢 Isolate ingestion kafka topics
    • 🟡 PersonHog
    • 🟡 Expand first point of capture to support better processing tooling
    • 🟡 Set ingestion SLAs and metrics and commit to them
    • 🟡 Improve per-partition ingestion throughput
    • 🟡 Make ingestion pipelines easy to manage

    Projects

    High priority

    • Isolate ingestion kafka topics
      • Clean up ingestion producers @pl
    • PersonHog @nickbest-ph @z0br0wn
      • Add client error observability
      • Restore latency alerts
      • Write somew runbooks
      • Investigate how to do person splitting in the PersonHog world
      • Django
        • Flag matching
        • Background tasks
        • Remove db connection from django-web and django-query
    • Expand first point of capture to support better processing tooling
      • Prepare the communication plan for rate limiting rollout @eli-r-ph @pl
      • Smoke test the v1 endpoint using the Rust SDK @eli-r-ph
      • Move capture-internal to v1 @eli-r-ph
      • Implement v1 in one more SDK @eli-r-ph
    • Set ingestion SLAs and metrics and commit to them
      • Break out the heatmaps pipeline @pl
      • Set up latency metrics for all pipelines @pl
      • Experiment with the SLOs dashboard @pl
      • Break out the ai pipeline for good @pl
      • Catch up with LLMA on the multimodal implementation @pl
    • Improve per-partition ingestion throughput
    • Deduplicator
    • Error tracking ingestion
      • Help with the v2 cymbal implementation @z0br0wn

    Incident follow-up

    • Look into deployments being stuck due to PVC issues
    • Incident review for 05.06

    Support follow-up

  20. jurajmajerik commented on Jun 9, 2026

    @jurajmajerik
    jurajmajerik
    on Jun 9, 2026
    Contributor

    Team Experiments

    ⚪ = Not Started 🟡 = In Progress 🟢 = Completed 🔴 = Won't complete 🔵 = Abandoned

    AI

    🟢 API layer - get the API ready for agents
    🟢 Skills - create experiment skills so PostHog AI knows how to use MCP tools effectively.
    🟢 MCP - supports all main operations.
    🔴 Signals - ship our first signal
    🟡 Feature flags - make it possible to run multiple experiments per feature flag
    🟡/🟢 Query performance - extend precomputation to the main subquery of all metric types.
    🟡 Loading experience - consolidate fetching and set better expectations in the UI.

    Retro

    Anders

    Marcel

    Rodrigo

    Juraj

    • Prompt experiments (hackathon project) now released as beta! We have ~5 companies using it, got some positive feedback and suggestions on what to improve.

    • Extended “Require completed conversion window” to retention metrics

    • Added a release mode choice when shipping variant - feedback from a customer. Confirmed with an experiment that it doesn’t introduce friction.

    • Timeseries fixes

      • Extend to 60-day experiments + warm the /query cache whenever a metric is calculated (-> main results available in the morning)

    • Working on a Slack agent that will proactively push updates to our Slack channel around our metrics/query performance/anything else. Good practice on writing and deploying my own agent.

    • Some UX fixes

      • Decluttering metric actions

    Ruby

    • RFC - End experiments with PR (closed)

    • RFC - for deployments (closed)

    • RFC - FF UX Review

    • Architecture and resource map - Feature Flags

    • RFC - Cost Margin Analysis (draft)

    Sprint Goals

    Anders

    • Support

      • Will also focus on papercuts, etc

    • Continue to push on getting feedback from feature flags on the RFC

    • Use the MCP extensively and iterate

    • Go through Github issues and clean up / fix stuff along the way

    Marcel

    • Continue with LLMA tooling

    • Come back to MCP

    • Extend MCP/Skills scope to further experiment resources/actions

    Rodrigo

    • Global on call June 1st to 7th

    • Release the metric reload temporal workflow.

    Juraj

    • Continue working on Prompt experiments as we’re getting good traction.

      • The Prompts UI is pretty rough, so will focus on that first.
      • Then continue adding more predefined metrics.

    • Continue working on the Slack agent

    • Decide where to go next with Signals x Experiments

      • For now I refocused on surfacing experiment ideas in PostHog Code onboarding step, since I don’t see a use case where Signals would be valuable for experiments at the moment.
      • Will start a thread with the Signals team to decide what to do here.

    Ruby

    • RFC - Cost Margin Analysis (draft)

    • Implement some suggested changes in RFC - FF UX Review

    • Follow-up from Architecture and resource map as now tagged

    • Entitlements convo

    • N8N - Experiments; largest by volume, but not in top 20

  21. marandaneto
    closed this as completedon Jun 15, 2026
  22. marandaneto
    unpinned this issue on Jun 15, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    sprintSprint planning

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions