AI Engineer (SaaS based company)

We are hiring for a SaaS based company and below are the details:

About Fortifai. Fortifai is an AI-native continuous risk intelligence platform. We detect anomalies and risk signals in procure-to-pay (P2P) and other financial processes by analysing enterprise ERP data (SAP and others) with modern ML and LLM systems. We are a Delaware-incorporated company with a wholly owned India engineering operation, building for enterprise finance, audit, and compliance teams. The three roles below support an aggressive build-and-scale phase across detection models, agentic pipelines, and customer deployment.

AI Engineer (Senior and Junior) - 2 positions

Location: India/Remote and onsite/Hybrid from Bangalore after a few months

Type: Full-time

We provide competitive salary

You will own the design and build of Fortifai's core detection and reasoning systems — the ML and LLM machinery that turns raw ERP data into trustworthy risk signals. This is a hands-on senior role with real architectural ownership and a path to growing a team under you.

What you'll do

• Design, build, and ship production anomaly-detection systems over large-scale financial and transactional data (time-series, statistical, and graph-based methods).
• Build LLM and agentic components — retrieval, grounding over structured data and knowledge graphs, multi-step reasoning, and natural-language explanation of findings.
• Own model evaluation and observability: build eval harnesses, golden-set regression tests, and the metrics that let us trust a model before it ships.
• Make the hard calls on when a problem needs statistical detection, classical ML, or an LLM — and when it doesn't need AI at all.
• Work directly with domain experts and product to translate finance/audit requirements into sound technical design.
• Mentor junior engineers; set engineering standards for the AI team.

What we're looking for

• 2 to 6+ years building and shipping production ML systems, with meaningful recent work on LLMs or GenAI (RAG, fine-tuning, agentic frameworks, evaluation).
• Depth in at least one of: anomaly detection, time-series modelling, graph ML / knowledge graphs, or NL-to-structured-data systems.
• Strong software engineering fundamentals — Python, clean production code, testing, data pipelines.
• Rigour about evaluation and failure modes; you measure before you claim, and you can explain why a model is wrong, not just that it works.
• Clear communicator who can reason about tradeoffs with both engineers and non-technical stakeholders.

Nice to have

• Experience with enterprise ERP or financial data (SAP, Oracle), or in fraud, risk, audit, or fintech.
• GraphRAG, Neo4j/Cypher, or structured-retrieval systems.
• Model fine-tuning / post-training (LoRA/QLoRA, SFT, DPO) and inference optimisation.
• Experience taking a system from prototype to a deployed, multi-customer product.

Background we value

We want engineers who have been trained by strong environments where real ML systems run at scale and engineering bars are high. Ideally your experience includes time at one or more of:

• A top-tier technology or product company, or a well-run AI-first company where ML is core to the product (not a support function).
• A global capability centre (GCC) or R&D arm of an established enterprise with mature engineering practices.
• A high-quality, well-funded startup (Series A and beyond) that has shipped real ML/LLM products to real customers.
• A recognised research group or lab, for candidates on the applied-research track.
• We care about the quality of the work and the environment, not brand names alone — a strong record at a serious startup counts for more than a title at a big company with little real ownership. What we're screening out is impressive-sounding experience with no production rigour behind it.