Skip to content
Search jobs inside your AI agent. Claude, ChatGPT, Claude Code, Codex, Gemini, Cursor and Grok
Connect

Senttr

NewBe an early applicant

Senior Software Engineer

Posted Updated

Summary

Senior engineer at Senttr building the platform behind its AI 'digital workers' — agents that automate admin like quotes, invoices and order entry for construction and manufacturing firms. Backend-heavy work in TypeScript/Node, Postgres and Kubernetes, plus LLM document understanding, tool-calling agents and system integrations.

This is a United Kingdom role only.


About Senttr

Senttr builds digital workers that take on the repetitive admin behind operational jobs. Our customers are operations-heavy businesses, mostly in construction and manufacturing, where much of the work is still done by hand. Each worker runs a whole job, with a person approving the steps that matter. Examples include handling enquiries, writing quotes and tender responses, creating sales orders, matching supplier invoices, checking H&S documents and chasing overdue payments.

Workers connect to the systems customers already use, such as Outlook, SharePoint, Xero, Sage, Business Central, HubSpot and Shopify.


The Role

We're hiring a senior software engineer to help build and run the platform behind our digital workers. You'll own features from design through to production across the platform.


What you'll work on

Our digital workers take on whole jobs for operations teams. They pick up what comes in by email, document or message, work out what needs doing, do it in the customer's own systems, and bring in a person when a decision matters. You'll help build the platform that makes this work:

  • Document understanding. Extracting structured, validated data from PDFs, scans, spreadsheets and emails using LLMs and OCR, and catching output that is well-formed but wrong.
  • AI in production. Tool-calling agents, structured output, model selection and fallbacks, and handling latency, throttling and model failure.
  • Workflow orchestration. Long-running, multi-step jobs that fan out across items, pause for human approval, retry and resume, with per-tenant concurrency limits.
  • Integrations. Building on third-party APIs such as Microsoft Graph, ERP, accounting and CRM systems: OAuth, rate limits, webhooks, and writing back to customer systems safely and idempotently.
  • Evaluation and observability. Tracing every run, building evals from real world data, and measuring accuracy from one release to the next.
  • Builder and chat. The web apps and agents customers use to create, configure and talk to their workers.
  • Multi-tenant SaaS Platform. Tenant isolation, safe releases across Kubernetes clusters/environments, security


How you'll work


We're a remote team, and as you'd expect LLMs play a large role in our day to day. We're believers in the technology, but we also don't like to throw slop grenades around. That means you need to be comfortable getting production grade work out of an agent, understanding where AI should not be playing a role, and also raising the bar generally on how we utilise our technology in the day to day. You'll be reporting to our CTO, but will maintain a high degree of autonomy and ability to influence our product.


What we're looking for


  • Curiosity and willingness to engage with problems, this is the most important aspect of being in our team.
  • Hopefully you have several years building and running production backend services, with ownership of features from design to deploy.
  • Daily, hands-on experience building software with AI coding agents (Claude Code, Codex, Cursor or similar): breaking work into tasks an agent can do well, writing specs it can follow, and steering it when it goes wrong. This doesn't have to be in a work setting.
  • The judgement to spot agent-written code that looks right but isn't, and the habit of proving a change against real data before calling it done.
  • Strong server-side TypeScript on Node.js, including a framework such as NestJS, enough to review it quickly as well as write it. Bun is a plus.
  • Solid SQL and relational data modelling. Postgres or SQL Server.
  • Experience shipping features built on LLMs: structured output, tool calling, prompts as code, and handling models that are slow, wrong or unavailable.
  • Comfort with containers and Kubernetes, and debugging a distributed system from logs and traces rather than guesses.


Nice to have


  • Rust in production (axum, tokio, sqlx).
  • Document extraction: PDFs, scanned forms, spreadsheets, OCR.
  • Integrating business systems: Microsoft Graph and Outlook, ERPs such as Business Central, accounting tools like Xero and Sage.
  • Evaluating AI output: building eval sets, model graders, measuring accuracy over time.
  • Running sandboxed or untrusted code safely.
  • Azure: AKS, API Management, Blob Storage, AI Search, Azure OpenAI.
  • Working with construction or manufacturing businesses.


Interview process


We generally have a 3 step interview process. An intro call with our CTO, where context will be shared about the role, some technical fundamentals will be discussed and a general "will we enjoy working together" check. A second technical interview that is a whiteboard system design exercise where you'll be expected to walk through an application idea and how we might build it together, followed by some deeper technical questions relevant to our product. Finally, a round meeting another founder to give you a chance to ask and answer any other outstanding questions about who we are and what we do.

Skills

What Senior Software Engineering jobs ask for — and how much of it you have →
Apply

See also

Software Engineering jobs by country — openings, pay and top skills →