Technical Lead Manager

Hevo Data Inc.; Hinoba-an; On-site; PHP 2,000,000 - 3,500,000; Full time

Job summary

Hevo Data Inc. is seeking an experienced backend engineering leader to shape and scale an AI-native data platform. You will guide architecture, set engineering standards, and own delivery across a growing team.

You will work with Java or JVM languages, Kafka and Kubernetes in production, mentor engineers, and own end-to-end reliability and multi-tenant considerations as the system handles massive data flows.

Qualifications

• 8+ years on backend or distributed systems.
• At least 1 year leading a team, official or informal.
• Experience running production systems with on-call, incidents, and SLAs.
• Strong Java or JVM language experience, with Kafka and Kubernetes in production.
• Experience with multi-tenant systems and high availability.
• Experience building with LLMs or AI integrations in production.

Responsibilities

• Lead architecture and build AI-enabled platform services and connectors at scale.
• Oversee pipelines from diverse data sources into Snowflake/BigQuery-like stacks.
• Ensure observability, reliability, security and auditability by default.
• Guide a team through production-grade delivery and on-call readiness.

Skills

Backend leadership

Distributed systems

Team mentoring

Java / JVM

On-call / incident management

Multi-tenant architectures

Tools

Kafka

Kubernetes

Job description

Read this first

You're probably already doing most of this job without the title. People bring designs to you before they build, your review is the one that catches things, and when something breaks late at night you're on the thread whether or not you're on-call. Juniors ask you about their careers even though nobody made that your job.

What we'd like to add is a team that's actually yours, a real say in who joins it, and the authority to decide how your part of the system gets built.

TLM is the step before Engineering Manager and we'd rather say that plainly than pretend otherwise. In 12 to 18 months you're either running a bigger team as an EM or you've worked out that you'd rather go deep, in which case Staff and Principal are open. Either way you won't be stuck.

You'd be running an AI-native team

• You will build the AI connector framework. The system that generates connectors, tests them, and repairs them on its own when a source system changes underneath.

• AI runs through every workflow, not just the product. Design, review, testing, production debugging. Teaching your team to work this way is part of the job.

• Real scale behind it. 100 billion records and petabytes of data every month, 2,000+ customers, and 150+ source systems that change their APIs without warning.

• The day job changes. Your engineers write specifications, evaluations and guardrails as much as application code, and you'll be judging output that isn't fully predictable. Few engineers in India are paid to do this yet.

What you'll actually be building

Platform services and database connectors that carry real weight. You'll lead the development of highly scalable, reliable and secure platform services and database connectors that power mission-critical data pipelines for thousands of enterprise customers. When a connector breaks, someone's revenue reporting is wrong that morning and they know it.

Pipelines into the modern data stack. These pipelines connect to warehouses like Snowflake, BigQuery and Databricks, as well as data lakes like Apache Iceberg. Each destination has its own semantics, its own failure behaviour and its own performance envelope, and your team owns getting all of them right.

Observability, reliability, security and auditability as defaults. These get embedded into everything your team ships rather than bolted on when a customer asks. You set that standard and you hold the line on it.

The hard problems underneath. Exactly-once delivery when any component can fail. Multi-tenancy where one customer's load never touches another's. Schema changes arriving mid-flight from a source system that gave no notice. None of these get solved once.

What's in it for you

You get to decide. Architecture in your area, what the framework does and doesn't do, where AI is the right tool and where it honestly isn't, and the hiring bar for your team. You won't be writing recommendations and waiting for someone else to approve them.

Your scope grows as fast as the company does. Engineering is up 30 percent in the last four to five months and that continues. Start with five engineers and you'll be running a considerably bigger team by the time you're done. That isn't on offer at a company standing still.

You get direct access to engineering leadership. No manager three levels down carrying your work upward. You'll be in the room when the calls get made, which is the fastest way to learn how to make them yourself.

You don't have to stop writing code. Roughly 60 percent of your week stays technical. Nobody here is going to quietly take the code away from you over four quarters.

You'll learn something most engineers can't yet. How to run a team where AI is embedded in the work rather than sitting next to it. In two years that will be expected of every engineering leader, and most people will pick it up late and secondhand.

What a good first year looks like

• The framework is generating production connectors, and the quality bar is set by evaluations you designed

• Shipping a new connector takes a fraction of the time it does today, and you can prove it with a number

• You've hired two or three engineers yourself and they're shipping

• At least one person on your team has been promoted

What we're looking for

• 8+ years on backend or distributed systems

• At least a year leading a team, and it doesn't have to have been official. If you're already the person everyone goes to, that counts

• You've run something in production and been there when it broke. On-call, incidents, SLAs

• Strong Java or another JVM language, with Kafka and Kubernetes in production

• Real experience with multi-tenant systems, high availability, or data at volume

• You've built something with LLMs rather than just used a coding assistant. Agents, evaluations, retrieval, prompt systems in production, any of it

• You can explain a complicated system to someone who isn't an engineer

• You've given someone hard feedback and kept the relationship afterwards

About Hevo

Pinterest, Gartner, Shopify, Iceland Air and Jawa Motorcycles run their data on Hevo, along with about two thousand other companies across 40+ countries. We move data from 150+ sources into Snowflake, BigQuery and Databricks without the customer writing code or putting an engineer on it.

Series B, $43 million raised from Sequoia Capital, Chiratae and Qualgro. San Francisco and Bangalore, with the platform built here.

We're not the biggest company you could join. We are one of the few where the engineering is hard, the team is growing fast enough that your scope grows with it, and the person building the system decides how it gets built.