Advanced types, JSONB, indexes (GIN/BRIN), partitioning, RLS, extensions, EXPLAIN.
Before starting, we run a 1-minute tech check — microphone, ambient noise, connection. If your setup isn't good enough, the test is fully refunded.
Show recruiters you actually know PostgreSQL — JSONB, GIN/BRIN indexes, partitioning, EXPLAIN, RLS — with a 15-minute AI-driven oral that goes way beyond a checkbox on your resume.
The Plume PostgreSQL badge certifies your ability to design, optimize, and operate PostgreSQL databases in real production environments. The 15-minute oral exam is run by an AI examiner that digs into your technical reasoning: why GIN over a B-tree? how do you read an EXPLAIN (ANALYZE, BUFFERS) plan? how do you pull off a zero-downtime migration using pg_repack or Flyway? The full transcript is then scored by Claude Opus, which produces a detailed report and a certified level from 0 to 100.
Unlike a multiple-choice quiz or a self-declared LinkedIn skill, this badge proves genuine understanding of Postgres internals: MVCC, the query planner's cost model, pg_statistic, Row-Level Security, declarative partitioning, logical replication, and extensions like pg_partman, PostGIS, or pgvector. The AI adapts its follow-up questions based on your answers, making every session unique and impossible to pass with surface-level knowledge.
This badge is built for backend and fullstack developers who rely on PostgreSQL every day, data engineers building pipelines on top of it, DBAs and SREs managing clusters in production, and anyone who wants to stand out in a market where 'knows SQL' has become meaningless noise. Whether you're gunning for a new role, a freelance contract, or a promotion, the PostgreSQL badge gives you a concrete, shareable signal where your resume used to leave a gap.
Here are the concrete dimensions the AI examines during the 15-minute oral.
Reading and interpreting EXPLAIN (ANALYZE, BUFFERS) output: spotting a costly Seq Scan, catching a bad join strategy (Hash Join vs Nested Loop), and diagnosing the gap between estimated rows and actual rows to find the root cause of slowness.
Picking the right index type for the job: GIN for JSONB and full-text search, BRIN for large append-only tables with physical correlation, GiST for PostGIS geometry or range types — and knowing when a partial or covering index is the smarter move.
Designing and managing range, list, and hash partitioning strategies — choosing the partition key wisely, ensuring partition pruning actually kicks in, automating lifecycle with pg_partman, and understanding the impact on VACUUM and query performance.
Modeling with JSONB: using operators like @>, ?, and #>>, building targeted GIN indexes on specific JSON paths, and making the deliberate call between semi-structured JSONB and a normalized relational schema depending on the access patterns.
Implementing RLS policies to enforce per-tenant data isolation in multi-tenant architectures, combining them with PostgreSQL roles, SECURITY DEFINER functions, and understanding where superuser bypass can silently break your security model.
Running schema changes on live production databases: adding nullable columns first, batched backfills, CREATE INDEX CONCURRENTLY, using pg_repack to reclaim bloat without a full table lock, and managing lock contention with lock_timeout and statement_timeout.
Fitting PostgreSQL into a real-world stack: choosing and configuring a connection pooler (pgBouncer, PgCat), wiring up migration tools (Flyway, Alembic), setting up logical replication to a read replica or data warehouse, and knowing how ORM-generated SQL can silently kill performance.
Knowing when Postgres isn't the right tool: massive OLAP workloads where ClickHouse dominates, ultra-flexible schema requirements where MongoDB makes sense, or horizontal write scale-out where CockroachDB or PlanetScale deserves a look.
Final scoring is performed by Claude (Anthropic), which reads back the full transcript and applies this weighted criteria grid.
Ability to diagnose a slow query using EXPLAIN (ANALYZE, BUFFERS), identify expensive nodes, choose the right fix between adding an index, rewriting the query, or refreshing statistics, and articulate the expected vs actual impact of the change.
Mastery of declarative partitioning, thoughtful use of JSONB vs relational columns, native PostgreSQL types (arrays, ranges, enums), and advanced constraints (EXCLUDE, deferrable FKs) that enforce data integrity at the database level.
Correct implementation of RLS policies, sound role and privilege management, solid understanding of VACUUM (standard vs FULL), autovacuum tuning, MVCC bloat mechanics — markers of someone who has actually run Postgres in production, not just queried it.
Quality of decisions around connection pooling, migration tooling, logical replication strategy, and extension selection (pgvector, PostGIS, pg_partman) — evaluated for coherence with real workload constraints, not just name-dropping.
Familiarity with PostgreSQL 15/16/17 features (MERGE, JSON_TABLE, improved logical replication, parallel query enhancements) and the ability to articulate where Postgres genuinely hits its limits compared to alternatives like ClickHouse or MongoDB.
A Plume session takes about 20 minutes, from tech check to badge delivery.
The AI runs a quick audio check to make sure your mic is working and you can hear the questions clearly. No tricks here — it's just a sound test before the real conversation starts.
You give a quick intro of your PostgreSQL background: your most recent or most complex project, the data volume involved, and what made it a genuine technical challenge. The AI uses this context to calibrate the depth of what comes next.
The AI works through the 8 calibrated themes: query optimization and EXPLAIN, index type selection, partitioning, JSONB modeling, RLS, zero-downtime migrations, stack integrations, and Postgres tradeoffs. It probes your answers with follow-ups to test whether your knowledge is real or rehearsed.
You can clarify any answer, mention a recent feature you've adopted in production (pgvector, MERGE, JSON_TABLE), or share what you're watching on the PostgreSQL roadmap. This is your chance to show genuine curiosity about the ecosystem.
Claude Opus reads the full transcript, generates a score from 0 to 100, assigns a certified level (Novice / Proficient / Advanced / Expert), and produces a detailed report broken down by evaluation dimension. Your PostgreSQL badge is live immediately with a shareable URL.
Your score out of 100 translates into a level a recruiter can grasp at a glance.
You use PostgreSQL like a slightly fancier MySQL: SELECT, INSERT, JOIN, and the occasional B-tree index. You've heard of JSONB or EXPLAIN but haven't used them in practice. You rely almost entirely on an ORM and can't explain the SQL it generates underneath.
You write complex queries (CTEs, window functions), use EXPLAIN ANALYZE to diagnose basic slowness, have created GIN indexes on JSONB columns, and manage migrations with Alembic or Flyway. You haven't yet ventured into partitioning, RLS, or MVCC internals.
You're comfortable with declarative partitioning, read EXPLAIN (BUFFERS) output with confidence, pick between B-tree, GIN, BRIN, and GiST for the right reasons, and have implemented RLS in a multi-tenant context. You understand the VACUUM/MVCC lifecycle and can run a zero-downtime migration using pg_repack or CONCURRENTLY.
You tune planner parameters (enable_seqscan, work_mem, random_page_cost), use targeted ANALYZE and pg_statistic to fix bad row estimates, and architect clusters with logical replication, pgBouncer, and automated failover. You run extensions like pgvector, PostGIS, or TimescaleDB in production and can articulate exactly when PostgreSQL isn't the right call.
No degree or years of experience required to take the badge. Here are the profiles it makes the most sense for.
You use PostgreSQL through an ORM every day but want to prove you understand what's happening under the hood — indexes, locks, execution plans — and that you won't freeze up when a query tanks in production.
You build pipelines that land and transform data in PostgreSQL. The badge validates your ability to model efficiently (partitioning, JSONB, native types), handle real data volume, and plug Postgres into a modern data stack without blowing up query times.
You run PostgreSQL clusters in production and want an external signal that backs up your expertise on VACUUM tuning, logical replication, RLS, and performance incidents — beyond what your job title conveys.
Clients always ask whether you 'know PostgreSQL'. The badge gives a scored, verifiable answer to that question and sets you apart from profiles who tick the SQL box without ever having read an EXPLAIN plan.
You're 1-2 years into working with Postgres and want to benchmark yourself honestly, get a detailed report on your gaps, and walk into your next technical interview with something more concrete than 'I've used it on a few projects'.
Where and how your PostgreSQL badge will help you day to day.
You're applying to a startup where PostgreSQL is the core database. Instead of a bullet point on your resume, you share your badge: score 78/100, Advanced level, with a report showing strong GIN indexing and partitioning skills. The recruiter has a real basis for a callback.
A potential client is deciding between you and another freelancer. You share your PostgreSQL Expert badge URL, backed by an oral where you explained how you cut a 4-second query down to 80 ms by combining a partial index with a CTE rewrite. The decision is made.
You're going for a lead data engineer role at your company. The badge lets you make your PostgreSQL growth tangible to your manager, with a dimension-by-dimension report showing exactly where you're strong and where there's room to push further.
You sit the badge two weeks before a technical interview. The report flags that your weakest area is reading EXPLAIN output with BUFFERS. Now you know exactly what to practice instead of reviewing everything at random — that's two weeks of focused prep.
You add your PostgreSQL badge URL to the Licenses and Certifications section on LinkedIn. Recruiters filtering on PostgreSQL see a scored, AI-evaluated badge rather than a self-declared skill — your profile stands out in the search results.
An engineering manager has the whole backend team sit the PostgreSQL badge to map real proficiency levels. Results show two Experts, three Proficients, and one Novice — giving a clear picture of where to focus code reviews, mentoring, and knowledge-sharing sessions.
A few minutes to check you have everything you need.
At the end of your session you don't just get a score — here's everything that awaits you.
Your PostgreSQL score and level (Novice / Proficient / Advanced / Expert) are computed by Claude Opus from the full transcript of your oral — not a quiz, but a real back-and-forth technical conversation.
You get a structured report across the 5 evaluation dimensions: query optimization, advanced modeling, security and administration, stack integrations, and technical awareness — with specific strengths and areas to develop.
The audio of your oral is stored securely and stays private to you. You can replay it to review your own answers and sharpen your preparation for your next PostgreSQL technical interview.
Your PostgreSQL badge gets a public URL you can add to your resume, LinkedIn, GitHub profile, or portfolio. Anyone who visits it sees your score, your level, and the context of the evaluation.
Discover related skills you can validate with Plume.
A 15-min oral exam with an AI, a shareable badge for your recruiters.
Choose this badge · €19.99
