AI Engineer

AI Engineer (Agentic Systems) at SureBright, Delhi, India

This is a high-ownership role for an individual seeking to operate at founder speed, gain comprehensive knowledge of the full stack of an insurance/warranty business, and deliver work that directly impacts revenue, conversion, and retention.

What You'll Do

You will be instrumental in building the agentic layer of our core product. This involves developing AI systems capable of reasoning, taking actions, and reliably completing workflows across critical business functions, including pricing, underwriting, policy issuance, claims intake, adjudication, and fulfillment (repair/replacement/reimbursement).

Key Responsibilities

• Design and deploy production-grade AI agents that manage real business processes, not just demos.
• Construct agentic architectures encompassing orchestration, tool calling, state machines, memory management, permission controls, audit trails, human-in-the-loop mechanisms, and fallback pathways.
• Take end-to-end ownership of our RAG (Retrieval-Augmented Generation) platform, including ingestion, chunking, embedding, retrieval, reranking, citations/grounding, and hallucination mitigation.
• Develop robust evaluation and monitoring systems, such as offline evaluation sets, regression tests, online metrics, drift detection, and red-teaming suites.
• Implement model optimization techniques, including prompt engineering, structured outputs, strategic fine-tuning, latency/cost optimization, caching, and throughput tuning.
• Build core ML systems for warranty and claims processing, focusing on document understanding, data extraction, classification, anomaly/fraud detection, decision support, and SLA routing.
• Collaborate closely with product and operations teams to translate real-world workflows into deterministic, testable, and compliant automation.

What You'll Build (Examples)

• Underwriting/pricing agents that provide real-time quote decisions based on merchant, product, and context signals, with stringent guardrails and auditability.
• A claims copilot and auto-adjudication engine for intake triage, evidence requests, decision proposals with explanations, vendor routing, and reimbursement automation.
• An OEM warranty parsing system designed to convert complex manufacturer policies into machine-readable coverage logic.
• Internal operations copilots to streamline manual work and enhance consistency across customer support, compliance, and finance.

Requirements (Must-Have)

(Hiring at different levels for the same role; required experience years and expected skill level will vary per role level)

• 1+ years of experience building and deploying ML/LLM systems in production, or equivalent founder-level experience.
• Proven experience in building agentic products or companies, including multi-step workflows, tool usage, orchestration, and reliability engineering.
• Deep hands-on expertise in:
  • RAG and retrieval systems (vector databases, reranking, grounding strategies).
  • LLM evaluation methodologies (golden sets, automated judging, human evaluation, regression pipelines).
  • Prompt engineering and structured outputs (schemas, function/tool calling, robustness).
  • Fundamentals and trade-offs of model training/fine-tuning (understanding when to tune vs. prompt vs. retrieve).
• Strong software engineering skills, including clean APIs, testing, observability, performance tuning, and secure-by-default design.
• Comfortable owning ambiguous problems end-to-end and driving them to measurable outcomes.

Strong Preference (Nice-to-Have)

• Experience building systems with compliance and audit requirements (e.g., in fintech, insurance, health, or enterprise sectors).
• Experience with document AI at scale, handling diverse inputs like PDFs, images, and messy data to reliably extract structured information.
• Experience designing human-in-the-loop workflows and escalation rules for high-stakes decision-making processes.
• Experience with infrastructure for LLMs, including model hosting, batching, streaming, caching, and prompt/version management.
• Startup or ex-founder background, particularly with experience shipping 0→1 products rapidly.

What Success Looks Like (First 90 Days)

• Ship an agentic workflow that significantly reduces manual operations work and improves a key metric (e.g., cycle time, accuracy, cost per claim, attach rate, CSAT).
• Implement an evaluation harness that effectively catches regressions before they reach production and provides a reliable quality score for each workflow.
• Establish a scalable architectural pattern for agents, including robust permissioning, audit logging, observability, and fallback mechanisms, that can be replicated by the team.

Tech Environment

Our environment is cloud-native and fast-paced. Expect to work with Python for ML/agents, TypeScript for product interfaces, Postgres for systems of record, event-driven services, and a modern LLM + retrieval stack with strong observability and CI/CD practices. Infrastructure is managed on AWS and Azure.

Why This Role Is Special

• Opportunity to build a category-defining, AI-native company in a massive market.
• Direct exposure to founders and high leverage, where your work will significantly shape the company's trajectory.
• Broad exposure across growth, underwriting/claims operations, and product development within a single role.
• A significant career accelerant; strong performance will lead to rapid growth in scope and title.

How To Apply

• Ensure your profile is up-to-date and includes a link to your LinkedIn profile.
• In your application message, provide three concise sentences, each describing something you've built or delivered along with its achieved results.