Software Engineer

ANRAK Corporation; Hyderabad; On-site; INR 3,200,000 - 6,000,000; Full time

Benefits offered by this job

Health cover for you and dependents

Latest hardware and AI tooling

Direct work with practising lawyers,Ar

Job summary

ANRAK Corporation in Hyderabad seeks an experienced software engineer to own meaningful parts of the product end to end, including retrieval and ranking over legal corpora, agentic workflows, and court-data pipelines. The role reports directly to the founder in a small team with no product-manager layer.

You will work on LLM pipelines, ingestion and normalization of court records, multi-step agent systems, and scaling a production stack with strict confidentiality.

Qualifications

• 4+ years shipping production software with Python and/or TypeScript.
• Experience with LLM applications: RAG, evals, prompt and tool orchestration.
• Strong backend fundamentals: data modelling, APIs, queues, observability.
• Ability to own ambiguous problems with little specification.

Responsibilities

• Own retrieval and ranking over legal corpora and agentic workflows.
• Build multi-step agent systems for case management, drafting, and hearings.
• Ingest and normalize court records, statutes, and judgments from Indian sources.
• Scale a production stack with confidentiality obligations and reliability standards.
• Shape engineering culture, tooling and hiring as the team grows.

Skills

Python

TypeScript

LLM apps

APIs

Observability

Data modelling

Ambiguity management

Tools

RAG tooling

Prompt orchestration

Monitoring

Job description

Anrak is an AI-native legal platform built for Indian lawyers - research across case law and statutes, court-record tracking, drafting and filing packs, paralegal workflow automation, and an institutional arbitration centre. We are moving from an India-first product to a Singapore-headquartered group with US and India operating companies.

The role

You will own meaningful parts of the product end to end: retrieval and ranking ove legal corpora, agentic workflows that paralegals and lawyers actually trust, court-data pipelines, and the infrastructure under all of it. Small team, direct line to the founder, no product-manager layer between you and the lawyers using what you ship.

What you'll work on

• LLM pipelines for legal research with citation verification and hallucination controls
• Ingestion and normalisation of court records, statutes and judgments from fragmented Indian sources
• Multi-step agent systems for case management, drafting, checklists and hearings
• Scaling a production stack serving law firms with real confidentiality obligations
• Shaping engineering culture, tooling and hiring as the team grows

You should have

• 4+ years shipping production software (Python and/or TypeScript; we care about depth, not stack loyalty)
• Hands-on experience with LLM applications: RAG, evals, prompt and tool orchestration, latency and cost control
• Strong backend fundamentals - data modelling, APIs, queues, observability
• Comfort owning ambiguous problems with little specification
• Strongly preferred: experience at a large-scale engineering organisation such as Google, Microsoft, Amazon or Meta, where you shipped systems at scale under real reliability and security standards
• Nice to have: legal-tech, search/IR, document processing, or regulated-industry experience

Compensation

We pay deliberately above market. We would rather hire three exceptional engineers at this rate than 6 average ones.

Where you'll work

Our office is at T-Hub, Hyderabad. This is an in-office role at the centre of Hyderabad's startup ecosystem, with the founding team on site.

Also

• Latest hardware and any AI tooling you want
• Health cover for you and dependents
• Direct work with practising lawyers, arbitrators and the founding team