Software Engineer job in Amman, Jordan, focused on AI agents. Full time, on site, with a fast growing YC backed AI startup building advanced AI agents for customer support. This is a backend heavy software engineering role for someone who has already built, shipped and optimised AI agents in production.
Mawahebna is running this hiring process on behalf of our client.
Job details at a glance
- Job title: Software Engineer, AI Agents. The internal title at the company is Customer Engineer
- Location: Amman, Jordan, on site
- Job type: Full time, permanent
- Experience: 3 or more years of hands on software engineering. Not a junior role
- Focus: Backend engineering with solid full stack ability, plus deep hands on AI agent work
- Core stack: Node.js, TypeScript, AI agents, LLM APIs, tool and function calling, RAG, prompt engineering
- Integrations: REST APIs, webhooks, and helpdesk and CRM systems such as Zendesk, Intercom and HubSpot
- Salary: 1,500 to 2,500 JOD per month
- Benefits: Health insurance
- Language: Fluent English, spoken and written. You will work directly with international customers
- Working hours: Amman based and on site, with some flexibility for customers across US, Europe and Saudi time zones
Read this first
This is first and foremost a hands on software engineering role. It is not a support role, not a research role, and not a front end role.
What matters most is strong backend engineering combined with real, extensive hands on AI agent work. You should be able to walk into an unfamiliar system, understand its architecture, trace a bug across services and third party integrations, and reason analytically about why an AI agent behaved the way it did, whether the cause sits in the prompt, the retrieved knowledge, a tool call or the data, and then fix it systematically. A machine learning research background is not required. Applied AI agent experience is.
You will work at the intersection of software engineering, AI agents, product and enterprise systems, with exposure to how global companies such as Booking.com and MoneyGram operate at scale. You will write production code, work deeply with AI agents, and dig into different businesses to understand their systems, workflows, integrations and customer journeys.
Your job is to understand the problem quickly, work out what is stopping an AI agent from performing at its best, and find the right technical solution. That could mean writing code, building an integration, improving how an agent uses tools or knowledge, debugging an API or a workflow, or taking an entirely different approach. There will not always be an obvious answer. We are looking for someone who enjoys figuring it out.
What you will do
- Build, improve and troubleshoot production AI agents and LLM powered systems
- Develop and ship production software in Node.js and TypeScript
- Learn how each customer’s systems and workflows actually work, then adapt the AI agents to perform inside them
- Diagnose complex problems and get to the root cause quickly
- Improve how agents use context, knowledge, tools, APIs and integrations
- Design and debug APIs, webhooks, integrations and agent workflows
- Work across backend and full stack systems when the problem needs it
- Come up with practical, sometimes out of the box solutions to problems that have no predefined answer
- Work directly with international customers and their technical teams
- Own problems end to end: understand, diagnose, design, build, validate
What we are looking for
Please apply if you strongly match the following:
- 3 or more years of hands on software engineering experience. Real production work
- Strong backend engineering with solid full stack ability. APIs, webhooks, queues, event driven systems and third party integrations. You trace a problem across services instead of stopping at the surface
- Strong production experience with Node.js and TypeScript. You write and ship real code
- Hands on experience building AI agents or LLM powered applications. You have built them, shipped them, and then had to make them work properly. Calling an LLM API once in a side project is not what we mean
- Solid understanding of LLMs, tool and function calling, RAG and retrieval, APIs, integrations and agent workflows. You know why an agent picks the wrong tool, and what to do about it
- Strong experience building and debugging APIs, integrations and production systems
- Excellent debugging and root cause analysis. When an agent answers wrongly, replies in the wrong language, or claims it did something it did not do, you can tell whether the cause is the prompt, the retrieved knowledge, the tool call or the data
- Ability to understand unfamiliar systems and business problems quickly
- Fluent English, spoken and written. You will be on calls with international customers and their technical teams, so this is not negotiable
- Confidence explaining technical ideas to engineers and non engineers alike
- High ownership, curiosity, and comfort in a fast moving startup. You prioritise speed, impact and getting things done
- Based in Amman, Jordan, or ready to relocate there, and comfortable working on site
Nice to have
- Experience with RAG pipelines, knowledge bases, or evaluating and improving the quality of LLM output
- Experience with agent frameworks, evals, tracing or observability for LLM systems
- Helpdesk and CRM APIs such as Zendesk, Intercom and HubSpot, or the ability to learn them fast
- Experience with AI products, customer communication platforms or enterprise SaaS
- Background integrating or extending messaging and communication systems
Who should apply
You do not need previous Customer Engineer experience. We are looking for strong software engineers with deep backend experience, solid full stack capability, and hands on experience building AI agents, who want to stay deeply technical while getting more exposure to AI, product thinking, enterprise systems and real customer problems.
If you have built with AI agents, love difficult technical problems, and enjoy working out a solution when the answer is not obvious, we would like to hear from you.
About the company
Our client is a fast growing, YC backed company building advanced AI agents for customer support, on an AI native customer experience platform used by enterprise businesses.
The product combines AI agents with deep integrations into helpdesk and CRM platforms, and handles customer conversations across chat, email, voice and messaging channels. Engineers work closely with customers, move quickly, and have direct impact on both the product and the customer experience. The company name is shared with candidates who move to the next stage.
Salary and compensation
The expected range is 1,500 to 2,500 JOD per month, depending on experience and overall fit, plus health insurance.
This role suits someone who enjoys being close to backend code, complex technical problems, AI agents and customers, and who wants to work in a fast moving startup environment in Amman.
Frequently asked questions
Is this a junior software engineering role?
No. You need at least three years of hands on software engineering, most of it backend, because a large part of the job is debugging complex issues in live customer environments. That instinct only comes from having built and broken real systems.
Is this a backend or a full stack software engineering job?
Backend first, full stack when the problem needs it. Your time goes into the Node.js and TypeScript backend, into integrations with customer systems, and into AI agent behaviour. Front end work happens, but it is not the focus.
How much AI agent experience do I actually need?
Enough that you have shipped something real and then had to make it behave. Building an agent, giving it tools, connecting it to a knowledge source, watching it fail in production and fixing it. If your only AI experience is a single LLM API call inside a side project, this role will be a stretch.
Do I need machine learning experience?
You do not need an ML research background and you will not be training models. What you do need is the ability to reason about an AI failure like an engineer. Expect to be asked in the interview how you would work through a real case: a customer reports that the agent has started replying in the wrong language, or that it confidently stated a policy that does not exist.
Is this Software Engineer role in Amman remote or on site?
On site in Amman, Jordan. You need to be based in Jordan and comfortable working from the office with the team. Amman hours are the base, with some flexibility when customers in the US, Europe or Saudi need it.
How good does my English need to be?
Fluent, spoken and written. You will be running technical conversations with international customers, so you need to be comfortable explaining a system, pushing back on a bad idea and writing a clear update, all in English.
Is this a customer support job?
No. It is a software engineering role with heavy customer exposure. You will read and write production code in Node.js and TypeScript, and ship fixes and improvements into the product yourself. The internal title is Customer Engineer, which is why we list it as a Software Engineer role: the work is engineering.
What is the salary for this Software Engineer job in Amman?
1,500 to 2,500 JOD per month, depending on experience and overall fit, plus health insurance.
Who is the employer?
A YC backed AI startup building AI agents for customer support. Mawahebna is running the hiring process on their behalf, and the company name is shared with candidates who advance to the next stage.
How do I apply?
Fill in the application form at the bottom of this page and attach your CV. You can also use the Apply Now button that follows you as you scroll.
Apply for this Software Engineer job in Amman
Send your CV through the form below. Every application is read by a real person at Mawahebna, and you will hear back either way.
