Vector search tuning can waste memory and hide recall regressions. This skill gives practical benchmarks, parameter guidance, and monitoring checks for production indexes.

Vector search projects need clear architecture choices before teams commit to an index or database. This skill guides retrieval design, embedding strategy, and tuning for production systems.

Teams need accurate Neon setup guidance across drivers, APIs, and local workflows. This skill directs Claude, Codex, and Claude Code to official Neon references.

Startup teams often track too many metrics or miss the numbers investors expect. This skill helps Claude, Codex, and Claude Code build focused metrics plans with formulas, benchmarks, and review cadences.

Startup founders need credible projections for planning, hiring, and fundraising. This skill structures assumptions into revenue, cost, headcount, cash flow, and scenario analysis.

Complex SQL work can slow teams when schemas, indexes, and workloads are unclear. This skill guides query tuning, schema design, and cloud database decisions.

Segment implementations often lose consistency across browser, server, and destination workflows. This skill gives Claude, Codex, and Claude Code structured guidance for tracking design and governance.

Segment tracking work requires precise schemas and identity rules. This skill guides Claude, Codex, and Claude Code through Rube MCP workflows for events, users, groups, pages, aliases, and batches.

Trading decisions can hide concentration, drawdown, and position sizing risk. This skill structures risk reviews, R-multiple tracking, hedging plans, and stop-loss discipline.

Portfolio risk reviews need consistent VaR, CVaR, drawdown, and risk-adjusted return calculations. This skill guides Claude, Codex, and Claude Code through clear risk metric patterns.

RAG projects often fail because retrieval quality is weak or unmeasured. This skill guides chunking, embeddings, vector search, hybrid retrieval, reranking, and evaluation choices.

Quant finance work needs disciplined modeling, realistic backtests, and risk-aware evaluation. This skill structures research with metrics, validation steps, and portfolio analysis guidance.

PostHog work often requires repeated lookup of projects, events, flags, and profiles. This skill guides Claude, Codex, and Claude Code through Rube MCP workflows.

Slow PostgreSQL systems need structured analysis before changes. This skill guides query tuning, index planning, configuration review, maintenance, and monitoring.

PostgreSQL schema design can fail when data types, constraints, indexes, and migrations lack workload context. This skill turns entities, access patterns, and scale targets into PostgreSQL-specific design guidance.

Slow queries and unsafe database patterns are hard to catch during ordinary code review. This skill gives Claude, Codex, and Claude Code structured Postgres guidance for indexes, RLS, pooling, transactions, and diagnostics.

Distributed NoSQL schemas fail when teams model entities before access patterns. This skill guides query-first design for Cassandra, DynamoDB, and ScyllaDB.

Production ML workflows often fail because data, training, validation, deployment, and monitoring are designed separately. This skill provides a structured lifecycle and orchestration guidance.

Manual Mixpanel analysis requires repeated project, query, and identifier setup. This skill guides structured analytics and profile workflows through current Rube MCP tools.

Single retrieval methods often miss exact terms or semantic matches. This skill helps combine vector and keyword results with practical fusion and reranking patterns.