We are hiring

Full-Stack Engineer

Own services end to end on a platform that runs real customer conversations at scale.

LOCATION

Bengaluru, Karnataka — on-site, full-time from office

EXPERIENCE

4–6 years

EMPLOYMENT

Full-time

REPORTS TO

Founder

CORE STACK

TypeScript / Node.js · Python · AWS · React / Next.js

COMPENSATION

Competitive, plus ESOPs. Discussed at offer stage.

APPLY TO [email protected]

ABOUT DESIBLE AI

Desible AI builds agentic conversation systems for banks, insurers and NBFCs in India. Our AI agents hold real conversations with real customers — over live phone calls today, across messaging channels next. We are in production with enterprise clients, in a regulated industry where security, privacy and auditability are requirements, not features.

We are a small engineering team. Everyone here owns what they build.

ABOUT THE ROLE

We are looking for a full-stack engineer to build and own services across our platform — APIs, backend services, the AI agent services behind our conversations, and the web applications on top of them. You will work on a multi-tenant, event-driven system that runs real customer interactions at scale, where correctness matters more than volume of code.

Our platform services are written in TypeScript and our AI agent work is in Python, so there is work here across both. Strong engineers come to us from either side — some with deep backend and product experience, some with more AI and agent experience. We are open to both. What matters is that you are a strong engineer who is willing to work across the stack; the rest is learnable here.

This is a hands-on role with end-to-end ownership: you will design, build, deploy and support what you ship.

RESPONSIBILITIES

- Design, build and maintain backend services and APIs in TypeScript / Node.js on AWS.
- Build and maintain AI agent services in Python — conversation logic, tool calling, structured outputs, and the integrations around them.
- Build and maintain web applications in React / Next.js on top of those APIs.
- Model data for the access patterns the product actually needs, and evolve schemas safely.
- Work with asynchronous, event-driven flows — queues, background jobs, scheduled work, retries — and make them reliable under failure.
- Write and maintain infrastructure as code for the services you own, and deploy them through CI/CD.




- Write tests that catch real problems: unit, integration and contract tests.
- Add logging, metrics and alerts so issues are found before a customer finds them, including tracing and observability for LLM-driven flows.
- Treat prompts and agent configuration as versioned, reviewed artefacts, and help build the evaluation that tells us whether a change made things better or worse.
- Build with security and data privacy in mind — access control, tenant isolation, secrets handling, minimal handling of personal data.
- Write explicit technical documents: specifications, design decisions and runbooks.
- Participate in code review and on-call for the services you own.

REQUIRED QUALIFICATIONS

- 4–6 years of professional software engineering experience, currently hands-on.
- Strong production experience with TypeScript and Node.js.
- Comfortable working in Python, or confident you can get there quickly — part of our work sits on the Python side.
- Solid database experience — relational or NoSQL — including schema design, indexing and query performance.
- Experience building asynchronous or event-driven systems: message queues, background processing, retries and failure handling.
- Working knowledge of AWS and experience deploying and running services in the cloud.
- Experience with infrastructure as code (AWS CDK, Terraform or similar) and CI/CD pipelines.
- Proficiency with React and Next.js.
- Practical experience with automated testing and with debugging issues in production.
- Sound understanding of application security fundamentals — authentication, authorisation, secure handling of credentials and personal data.
- Clear written and spoken English.
- Able to work full-time from our Bengaluru office.

GOOD TO HAVE

None of these are required. They are the areas we would be glad to find, and any one of them is worth mentioning in your application.

- Hands-on experience building with LLM APIs — prompting, structured outputs, function / tool calling,



and handling non-deterministic output sensibly.
- Experience with multi-tenant SaaS products.
- Exposure to security and compliance processes such as VAPT, SOC 2, ISO 27001 or India's DPDP Act.
- Experience with search or analytics systems, or with reporting and dashboard products.
- Experience with telephony, real-time or streaming systems.
- Experience with agent orchestration frameworks such as LangGraph, or with hand-built agent state machines.
- Experience with LLM observability and tracing tools such as Langfuse, LangSmith or OpenTelemetry-based setups.
- Experience building evaluation for LLM systems — test sets, LLM-as-judge, regression checks.
- Experience with retrieval-augmented generation, or with speech systems (STT / TTS).
- Familiarity with AI coding assistants (Claude Code, Cursor, Copilot) as part of your daily workflow.

WHAT WE OFFER

- Direct ownership of services, end to end, with no layers in between.
- Working directly with the founder on product and technical decisions.
- Competitive compensation with ESOPs.
- A live product with enterprise clients — real users, real scale, real feedback.
- Fast development space: A high-velocity environment where we prioritize rapid iteration and deployment.
- Spec-level development: A mature engineering culture where we build against clear, defined specifications to ensure alignment and quality.

HOW TO APPLY

Email your CV to [email protected] with the subject line “Full-Stack Engineer — ”.

Please include short answers to the following in your email:

1. Describe one system you built and owned in production. What was the hardest problem you solved in it?

2. Have you worked on anything that processed work from a queue or ran on a schedule? What did you do to make sure the same work was not processed twice?

3. If you have built anything with LLMs — a production feature or a serious side project — tell us briefly what it did and what you found hard about it. Skip this if it does not apply.

4. A link to code you are proud of, with a line on why.

5. Your notice period, current and expected compensation, and confirmation that you can work from our Bengaluru office.

We review every application that includes these answers and respond either way.

desible.ai · [email protected] ·

-

📌 Software Engineer (Bengaluru)
🏢 Desible.ai
📍 Bengaluru