Senior Backend Engineer

Byggveai Technology AB; Stockholms kommun; On-site; SEK 900,000 - 1,200,000; Full time

Job summary

Brickanta, an AI-focused company, is building a robust backend that ingests documents, specs, and drawings, turning them into structured data for AI reasoning. You will own core services and data pipelines, shaping how the product processes inputs at scale.

We value builders who ship over credentials, expect you to work in Sweden, communicate in English, and be on call for reliability. You will influence data models, service boundaries, and observability as the system grows.

Qualifications

• Shipped and operated backend systems in production.
• On-call experience and debugging at 2am.
• Strong understanding of data flows in real-world data.
• Ability to make architectural decisions that scale.

Responsibilities

• Own core backend surfaces end to end.
• Design data models, APIs, and pipelines for fast, reliable processing.
• Ensure observability, reliability, and performance under load.
• Collaborate with customers and the team to translate requirements into robust solutions.

Skills

SQL

Backend development

System design

On-call experience

Tools

Python

FastAPI

TypeScript/Node

PostgreSQL

Redis

AWS

Docker

Job description

About the role

Brickanta is built by an AI-veteran team. We've been shipping AI systems since 2018 and have built technical products inside some of the world's biggest brands. Now we're pointing that at a $15 trillion industry that software mostly forgot. We build for a domain where reliability matters: high-stakes decisions, messy inputs, real money on the line.

We're 20 people and moving fast. At our size you own big parts of the backend that the whole product stands on. You'll get fuzzy problems like "document ingestion needs to be fast and reliable," along with the room to decide what to build and the mandate to ship it.

We don't optimize for credentials. We optimize for builders who ship.

What you'll build and own

• The core services behind bid analysis, estimation, and procurement: APIs, data pipelines, background jobs, and the plumbing that turns messy construction inputs (documents, specs, standards, drawings) into structured data our AI can reason over and our customers can trust
• The data layer itself. Schema design, query performance, and the indexing and batching decisions that keep things fast as records pile up
• The architecture that outlives any single feature. You make the calls on data model, service boundaries, when to add a queue, and what breaks at 10x, and the rest of the team builds on them for years
• Reliability and evaluation infrastructure: the systems that measure quality, catch regressions, and make AI output repeatable and measurable
• Performance and observability that hold up as usage scales, so the product stays fast and debuggable under real load
• Integrations and forward-deployed problem-solving that make it work in the customer\'s reality
• The systems live in production, which means you run them. You\'re on call for what you build, and you write code that\'s debuggable when something goes wrong at 3am

You\'re probably a fit if you

• Have shipped and operated backend systems in production. You\'ve been on call, debugged the thing at 2am, and learned from it
• Have real fluency with SQL and relational data. You anticipate bottlenecks and can read the metrics after you ship to confirm it\'s working
• Have the judgment to know when to build something properly versus ship a rough version and revisit, which corners are safe to cut, and when "clever" is a liability. You\'ve felt the consequences of getting this wrong
• Reason from user pain → solution → measurable outcome
• Care about correctness in messy, real-world data, including the edge cases that only appear in production
• Move fast and care about craft at the same time
• Raise the people around you through code review, patterns others follow, and mentoring, without slowing your own shipping
• Work with little management. You surface risks early, push back on shaky requirements, and translate between technical reality and business need, including talking to construction customers directly
• Have opinions about AI and LLM systems in production (evals, reliability, latency, cost) or are hungry to develop them

How we work

AI is core to how we build, not a bolt-on. We work AI-first day to day, encode our conventions and hard-won lessons into skill files, and expect everyone here to sharpen that craft as agentic engineering matures. If you\'ve been pushing the edge of how AI fits into real engineering work, you\'ll fit right in.

What success looks like

• In your first month you\'ve shipped something real to production, you understand our stack and the shape of construction data, and you\'re already flagging what\'s fragile
• By your first quarter you own a core backend surface end to end. When it breaks, people look to you, and it breaks less because of decisions you\'ve made
• By your first year the architecture you drove is load-bearing across the product, new engineers ramp faster on the patterns you set, and the systems you own are ones the team trusts without thinking about them

Tech stack

Something like: language and framework (e.g. Python/FastAPI, TypeScript/Node), primary database (e.g. Postgres), queue and cache (e.g. Redis), infra and deployment (e.g. AWS, Docker), plus the LLM tooling you use. List what a hire will actually touch. You can add the standard Brickanta line: prior experience with our exact stack is a nice-to-have, not a requirement.

How we evaluate

• Show us what you\'ve built: repos, demos, shipped products, systems you\'ve owned end to end
• What you\'ve made work counts for more than where you worked

Note: Sweden is Brickanta\'s first and most important market. Working comfort in English is required; Swedish is a plus but not required.

Keen on doing your life\'s work, bringing one of the world\'s largest and most impactful industries, the $15 trillion construction industry, into the AI era?