Why this role exists

We're building AtomsAI to deliver AI transformation in weeks, not quarters  powered by our own proprietary AI operations platform for workflows, knowledge, reviews and actions. The platform is deep underneath (RAG, agents, ontology-driven business objects, model gateway, eval and replay, private VPC deployments) and deliberately simple on the surface.

We need engineers who can take a scoped customer problem and ship a production-grade AI system end-to-end and then keep it working when real users, real data and real edge cases show up.

What you'll actually do

• Build RAG pipelines and agentic workflows end-to-end - retrieval, context assembly, tool use, guardrails, evaluation.

• Orchestrate LLM systems with LangChain / LlamaIndex / LangGraph (or equivalent) - and know when to drop the framework and write it yourself.

• Integrate across enterprise APIs, webhooks and comms platforms (CRMs, helpdesks, telephony, document systems).

• Debug production AI: latency, retrieval quality, drift, weird model behavior, flaky integrations. Own the pager when it matters.

• Contribute back to shared platform tooling - recipes, evals, connectors - so the next engagement is faster than this one.

• Work directly with customer engineering and ops stakeholders. You'll be in the room, not behind a ticket queue.

What we're looking for

• 4–6 years of software engineering experience, with at least 2 years shipping LLM-powered systems in production (not just notebooks or demos).

• Strong hands-on RAG: vector stores (Pinecone, Weaviate, pgvector, Qdrant), embedding models, retrieval strategies and how to evaluate them honestly.

• Proficiency with LLM orchestration frameworks and the judgement to know their limits.

• Experience building agentic systems - tool use, multi-step reasoning, memory, guardrails - that survive contact with real users.

• Strong Python, comfort with REST, webhooks, async and queue-based systems.

• Prompt engineering as a craft, not a vibe - structured prompts, few-shot, CoT, versioning across environments.

• Ability to debug AI in production. Latency budgets, retrieval quality regressions, model-behavior changes - you've fought all three.

• Direct communication with technical and non-technical stakeholders.

Edges that help

• Hyper-growth or high-scale environments.

• Cloud infra, deployment and production debugging (GCP especially).

• Enterprise integration pipelines and automation at scale.

How we work

• Outcomes over hours. Impact is the metric.

• 5 days in office - intensity, team-work & solving tough problems

• Annual offsite & numerous fun parties

• Comprehensive health coverage for you and your family.

• Competitive pay tied to skill, experience and the value you ship.

The bigger picture

AtomsAI sits inside SaaS Labs - the team behind JustCall, CallPage and ServiceAgent.ai, with 6,000+ customers, 300+ people across India, Poland and the US, backed by Peak XV and Base10 Partners (raised over $55mil). AtomsAI is our newest bet and the engineers we hire now will shape the platform thousands of customers will eventually run on.

If you want to build AI that actually works in production - for real users & real business problems reliably - let's talk.

SaaS Labs takes pride in being an equal opportunity employer. We believe diverse teams build better products, which is why we're committed to taking hiring or any employment decision regardless of race, color, religion, national origin, age, sex, marital status, ancestry, disability, genetic information, veteran status, gender identity, sexual orientation, or any other characteristic.

Curious about what we're building? Follow us on LinkedIn to be the first to know!