Website:
revmozi.com
Job details:
Location: Bengaluru, in office
Experience: Typically 2–4 years
Employment: Full-time
About RevMozi
RevMozi is building an always-on AI GTM operator for early-stage B2B SaaS companies.
Our agent continuously studies a company and its market, identifies opportunities, prepares campaigns and content, coordinates approvals, executes work across external systems, and learns from the results.
This is not a prompt box with an LLM attached. We are building an agent that operates over long-running workflows, persistent company knowledge, tools, approvals, and real-world side effects.
The role
We are looking for a product-minded engineer who can own systems end to end - from agent execution and backend services to the interface through which users supervise the agent.
You will work across three connected layers:
- The AI-agent runtime: context, tools, memory, workflows, evaluations, permissions, and failure recovery.
- The platform: APIs, data models, integrations, authentication, background jobs, and observability.
- The product: clear, responsive interfaces that help users understand, approve, and control the agent’s work.
Backend engineering will be your centre of gravity, but this is not a backend-only role. You should be comfortable following a user problem through every layer required to solve it properly.
What you will own
- Build and operate production agent workflows that can run reliably over minutes, hours, and days.
- Design tools and interfaces through which agents safely read data and perform actions.
- Build APIs, relational data models, asynchronous jobs, and third-party integrations.
- Develop product interfaces in React and Next.js when the work crosses into the user experience.
- Design approval, permission, retry, idempotency, and failure-recovery behaviour for consequential agent actions.
- Instrument systems so production journeys can be traced and failures diagnosed quickly.
- Work directly with the founders to turn ambiguous customer problems into simple product behaviour.
- Review your own work in production and improve it based on evidence rather than merely closing tickets.
What we are looking for
- 2 to 4 years of experience building and operating production software or equivalent evidence of ability.
- Strong backend fundamentals: APIs, SQL, relational modelling, asynchronous execution, authentication, and failure handling.
- Proficiency in TypeScript or demonstrated ability to become productive in it quickly.
- Experience shipping credible frontend work using React or a comparable framework.
- Experience building with LLMs in production or a substantial project demonstrating agentic system design. Prompt engineering alone does not count.
- The ability to debug a problem across application code, infrastructure, data, and external providers.
- Good product judgment: you can distinguish what a user needs from what is merely technically interesting.
- Clear written communication and the ability to make progress without a detailed specification.
- A record of owning outcomes rather than only implementing assigned tasks.
Particularly valuable experience
You do not need every item below:
- Tool-calling agents, MCP, context engineering, evaluations, or agent memory.
- Durable workflows, queues, scheduled jobs, webhooks, and idempotent processing.
- PostgreSQL and schema design.
- Next.js, React, Hono, Drizzle, Bun, or Cloudflare.
- OAuth and integrations with third-party SaaS platforms.
- Observability for multi-step or distributed workflows.
- Building an early-stage product where requirements changed as the team learned.
What this role is not
- A narrow API implementation role.
- An ML research or model-training position.
- A collection of prompt-writing tasks.
- A job where product managers hand you fully specified tickets.
- A role for someone who wants to remain exclusively on either the frontend or backend.
What success looks like
Within your first 2 weeks, you have shipped a meaningful improvement across at least two layers of the system.
Within 1 month, you independently own a user journey- from agent decision through execution and user visibility- and can diagnose its production failures.
Within 3 months, you have materially improved either the agent’s capability, the reliability of its execution, or the speed at which the team can safely ship new workflows.
How we work
- We work together from our Bengaluru office.
- Engineers participate in product decisions and speak directly with users.
- Ownership includes design, implementation, deployment, observation, and iteration.
- We prefer small, legible systems over speculative abstractions.
- We use AI heavily, but we do not confuse generated code with engineering judgment.
- We care about pace and quality. Neither excuses the absence of the other.
Applying
Send us:
- A short introduction and links to your GitHub, portfolio, or work you can discuss.
- One product or system you owned substantially.
- The hardest production problem you personally diagnosed.
- An example of something you built using LLMs or agents, if applicable.
- We care more about what you can demonstrate than where you studied.
- Email them at talent@revmozi.com
Click on Apply to know more.