Open Position
Software Engineer
Dhaka, Bangladesh
Full-Time
On-site
2-4 Years
Nagorik Technologies is an AI-native enterprise technology company with 200+ in-house engineers building software, AI systems, cloud platforms and secure infrastructure for complex, business-critical operations. We have built, run, licensed and white-labeled our own products, so we own what happens after launch: reliability, security, scale and maintainability. Our work spans fintech, telecom, SaaS, healthcare, government and retail. We are ISO 9001:2015 and ISO 27001:2022 certified and an AWS Partner.
How we build software now
Our engineers use AI agents across the software lifecycle, from analysis and design to coding, testing and operations. Agents do not replace engineering judgment. They raise the bar for it, because an agent will confidently do exactly what a vague instruction allows. The skills that decide the result are:
- Context: understanding the business problem, the existing system, the data and the constraints well enough to give an agent what it needs, and to know when its output is wrong
- Specification: writing clear, testable specs and acceptance criteria that leave nothing to guess
- Prompting: writing precise, structured instructions, examples and project rules for the task at hand
- Loop engineering: designing the plan, act, verify and retry cycles an agent runs, including when it must stop and ask a human
- Harness engineering: building the tools, permissions, sandboxes, tests, evaluations, hooks and CI gates that keep agent work inside safe, measurable limits
- Verification and ownership: reviewing every change, proving it works, and being accountable in production
We measure engineers on outcomes shipped, defects avoided and systems that stay up, not on lines typed. A Claude certification is an advantage at every level.
The role
As a Software Engineer, you own features from the first conversation with a product owner to the moment they run reliably in production. You use AI agents every day, and you know that their output is only as good as what you put in: the context you gather, the spec you write, the prompt you craft and the checks you build around them. You decide what gets built and how it will be proven correct, you direct agents to build it, and you stand behind the result after release. You will work on APIs, integrations and data platforms that serve millions of users and high transaction volumes.
What you'll do
- Own features end to end: understand the business need and the surrounding system first, then design the approach, build, verify, ship and monitor
- Write specs, interfaces and acceptance tests precise enough that an agent cannot reasonably misread them
- Write and refine prompts, project rules and examples, and know which context an agent needs for each task
- Run and tune agent loops: plan, implement, test, fix, repeat, with clear stop conditions and points where a human decides
- Contribute to our harness: test suites, linters, hooks, CI checks and evaluations that catch bad output automatically
- Review agent-generated changes critically for correctness, security, performance and maintainability
- Build and maintain application APIs and data access layers, including third-party REST and SOAP integrations
- Investigate production issues, find root causes, and turn each lesson into a better spec, test or rule
- Prototype new tools and techniques and share what you learn with the team
What you bring
- Bachelor's degree in CSE or a related field
- 2 to 4 years of professional experience as a backend or full-stack engineer
- Working experience in Laravel/PHP or Node.js with NestJS, and in JavaScript/TypeScript
- Strong understanding of OOP, data structures, relational databases, MVC and API design
- Experience integrating third-party REST or SOAP APIs, and designing reliable queues and background jobs
- Practical experience with AI agents on real work, and the habit of diagnosing a bad result: missing context, a vague spec, or no feedback signal
- Clear written communication: you can write a spec another engineer or an agent can implement without asking you questions
- Confident with Git, code review and CI/CD
- Ability to read and work in code written in languages or frameworks you do not know
- Self-driven learning and a sense of ownership
Advantages
- A Claude certification
- Experience building agent loops or harnesses: tool definitions, permissions, automated checks or evaluations
- Next.js, Vue.js, Python or Go
- Cloud experience with AWS, DigitalOcean or Google Cloud, plus Docker
- Exposure to payments, fintech or telecom integrations
What you will grow into
You will take on larger, riskier features, begin mentoring juniors, and shape how our team builds its loops and harness. The path leads to Senior Software Engineer.
What you get
- Competitive salary according to experience: [BDT range to be confirmed]
- Two days weekly leave
- Two festival bonuses
- Yearly performance-based increment
- Health insurance and wellness programs
- Complimentary breakfast, lunch and evening snacks
- Flexible working hours
- Fast growth in a company that is expanding its engineering team
- Real impact: you see how your work reaches users
How we hire
Screening call, then an exercise that starts from a vague requirement. We watch how you ask questions, gather context, write the spec, direct AI tools and verify the result, and then ask you to walk us through your decisions. Next comes a technical discussion about a system you have built, and a final conversation with the team.