Backend audit · fixed scope · 2 weeks
Find out where your backend is slow, fragile or expensive.
A measured picture of a Kotlin or Java backend and a prioritized plan to fix it, from the engineer who took search over 25–50M records from 30+ minutes to under a second and built a crash-safe export pipeline that never loses a record. Decisions rest on numbers, not guesses.
What the audit covers
Performance
Slow queries, missing or wrong indexes, pagination that degrades with data volume, N+1 access, hot paths that block the request thread. Measured on realistic data, not guessed.
Reliability
Background jobs that can be lost, run twice or get stuck; retries without idempotency; consumers that break on duplicate delivery; failure modes that surface as lost orders or support tickets.
Observability
Whether one request can be followed across services and brokers, whether logs, traces and metrics connect, and whether alerts page on what matters to the business.
Architecture & cost
Service boundaries, data ownership, coupling and infrastructure spend. Where the system is more complex or more expensive than the problem requires.
How it runs
01
Intro call · 30 minutes
The business problem, the current system, what access is possible. We agree scope and price after this call.
02
Days 1–7 · Read and measure
Read-only access to code, schemas and production metrics or a staging copy. Benchmarks on realistic data volumes.
03
Days 8–10 · Report
A written report: each finding with evidence, business impact, effort estimate and a recommended order of work.
04
Walkthrough call
I walk your team through the findings and answer questions. The fixes can be done by your team, or by me as a separate fixed-scope project.
A good fit if
- Your backend is Kotlin or Java (Spring Boot), with PostgreSQL, MongoDB, Oracle, Kafka or RabbitMQ behind it.
- The product is growing and the system is getting slower, less predictable or harder to debug.
- You want a measured, prioritized picture before spending on new infrastructure, licenses or a rewrite.
Proof of approach: search at 25–50M records,a crash-safe async pipeline andone trace ID across 7 services.
Scope and price are agreed after the first call. NDA on request. B2B invoice in USD or EUR.