Software Engineer - AI
Job type: Full Time · Department: Technology · Work type: On-Site
Bengaluru, Karnataka, India
ACKO is the protection destination for over 200 million tech-savvy families across India, protecting their families, assets and money. Launched in 2016, ACKO started by reimagining insurance, making it simple, hassle-free and customer-first. Today, our mission goes beyond that: we aim to touch the lives of 1 million users, building products that solve real-world problems with technology at the core.
We are not just another insurance company, our DNA is product-tech, and our approach is bold, innovative, and digital-first. From zero commission, zero paperwork, and instant renewals to same-day claims settlements and app-based tracking, ACKO is a Welcome Change from traditional insurers.
But what truly sets us apart? Our people. At ACKO, every Acker’s voice and ideas matter. We’re a vibrant, inclusive team of creators, thinkers, and doers, building products that redefine protection while ensuring each Acker grows, thrives and does meaningful work.
Join us at ACKO, where bold ideas, real impact and tech-driven innovation redefine protection and peace of mind - and where YOU can make a real difference in people's lives. ACKO is a product-tech company, launched in 2016, solving real-world problems for customers, starting with insurance. And as a customer-first organization serving the digitally-savvy, ACKO’s value proposition of ‘Welcome Change’ focuses on offerings that make insurance simple and hassle-free! With features such as zero commission, zero paperwork, instant renewal, same-day claim settlements, and app-based updates on claims, ACKO is a 'Welcome Change' from traditional insurers.
Having said that, we are not just another conventional insurance firm, or the people consulted solely for "claims”! Anchored in a tech-centric philosophy, ACKO’s approach fuels innovation, empowering us to develop comprehensive products that cater to every aspect of our customers' insurance requirements. And while we are at it, we put our Ackers at the heart of everything we do. We're not your typical 9-to-5 workplace; we're a vibrant and inclusive bunch of innovators and creators making sure every Acker’s idea matters, their voice is heard, and their growth is part of our mission.
Insurance is fundamentally a data and decision-making business. ACKO's D2C model means we talk to millions of customers directly - policy discovery, underwriting signals, claims guidance, renewal nudges. Every one of these touchpoints is being redefined by AI.
This engineer won't be a data scientist. But they must understand enough about LLMs, inference pipelines, and model behaviour to make sound engineering decisions: when to call a model vs. a rules engine, how to prompt reliably at scale, how to build fallback logic, and how to measure whether an AI feature is actually working.
Concretely, AI capability in this role means
Designing backend systems where LLM calls are first-class, not bolted on - rate limiting, prompt versioning, cost budgets, output validation
Owning the integration layer with AI services (OpenAI, Bedrock, etc.) including retries, caching, abuse prevention, and observability
Building AI-enabled product features: intelligent document processing, conversational flows, automated claims triage, smart nudges
Making the build-vs-buy-vs-prompt call - and justifying it with data
Evaluating LLM outputs in production: building evals, red-teaming prompts, and catching regressions before they hit customers
The Non-Negotiables
3-8 years of backend or full-stack engineering experience on large-scale, production systems
Strong CS fundamentals: data structures, distributed systems, concurrency, system design
Fluent in at least one high-level language; Java/Spring preferred but not exclusive
Hands-on experience integrating with AI/ML services or LLM APIs (OpenAI, AWS, Bedrock, Cohere, Anthropic, or similar) in production
Demonstrated ability to design backend systems that support AI-driven workloads: async inference, streaming, structured output parsing, fallback logic
Working knowledge of prompt engineering - not just "write a prompt" but version, evaluate, and harden prompts for reliability
Experience with databases (SQL + NoSQL), caching, queuing — the fundamentals haven't gone away
Strong debugging and observability instincts: traces, logs, metrics, alerts
You've shipped an AI-powered feature end-to-end and can speak to what broke, what you measured, and how you iterated
You have a view on when NOT to use an LLM - and you can defend it
Experience with GenAI application patterns: RAG, tool use / function calling, multi-step agents, structured output
Familiarity with vector databases, embeddings, or semantic search (pgvector, Pinecone, Weaviate, or similar)
Experience deploying or monitoring AI models in production: latency budgets, cost tracking, output evals
Exposure to AI-assisted developer workflows and the ability to multiply your own output using these tools
Mobile-aware mindset — understanding how AI features surface in a mobile-first product
Security and abuse prevention instincts applied to AI: prompt injection, output filtering, rate limiting
We're building a team that doesn't just adopt AI - it shapes how AI is used responsibly, reliably, and meaningfully in financial services. If that sounds like the problem space you want to work in, we'd like to talk.
Autofill application
Save time by importing your resume in one of the following formats: .pdf or .docx.