Senior AI Forward Deployed Engineer

Fractal Analytics Private Limited; Bengaluru; On-site; INR 3,500,000 - 5,200,000; Full time

Job summary

Fractal Analytics Private Limited in Bengaluru seeks a Senior AI Forward Deployed Engineer for Cogentiq I2C to own technical outcomes across a portfolio of accounts. You will translate finance problems into production AI systems, lead a deployment team, and set KPIs like accuracy, latency, and business impact.

You will design architectures, own integration with ERP, banking, and document systems, and mentor engineers while guiding governance and risk considerations.

Qualifications

• Strong Python and system design depth.
• Depth across LLM and agentic architectures, retrieval, evaluation methodology, guardrails and observability.
• Cloud deployment experience at scale; Azure preferred.
• Containers and orchestration (Docker, Kubernetes).
• Experience mentoring or leading engineers.
• ERP integration experience with SAP S/4HANA or Dynamics 365 Finance and Operations.

Responsibilities

• Own the technical outcome for a portfolio of accounts through production rollout and scale.
• Translate goals into solution design covering topology, retrieval strategy, and data model.
• Lead a deployment team and define effort estimates and success criteria.
• Explain system capabilities and residual risk to leadership; guide governance and model risk.
• Convert deployment learnings into reusable assets and feed the product roadmap.

Skills

Python

System design

LLM architectures

Agentic architectures

Retrieval & eval

Cloud deployment

Mentoring engineers

Tools

Azure

Docker

Kubernetes

SAP S/4HANA

Dynamics 365 Finance

Job description

Senior AI Forward Deployed Engineer - Cogentiq I2C

It's fun to work in a company where people truly BELIEVE in what they are doing! We're committed to bringing passion and customer focus to the business.

Role summary

You own the technical success of Cogentiq I2C at a portfolio of strategic accounts. You take a business problem that arrives stated in finance terms, reduce our days sales outstanding, stop writing off small-balance deductions, and turn it into a production AI system that the customer's finance organisation actually uses. You lead a small deployment team. You make the architecture calls. You are the person the customer sponsor calls when something is wrong, and the person who tells them what we will not build. This is an engineering role. You will still be in the codebase. The difference from the first band is that you also carry the account, the design and the team.

Account technical ownership

Own the technical outcome for a portfolio of accounts, from discovery through production rollout and scale. Set and defend the deployment KPI contract: accuracy, straight-through processing rate, latency, cost per transaction, adoption and business impact. Hold the technical relationship with customer sponsors and IT leadership.

Solution design and architecture

Translate ambiguous goals into a solution design covering agent topology, retrieval strategy, human oversight boundaries, integration approach and data model. Make and document the build versus configure versus extend-the-product calls. Decline bespoke work that cannot be maintained across the customer base. Own the integration design against customer ERP, banking and document infrastructure.

Delivery leadership

Lead a deployment team of Forward Deployed Engineers and specialists. Scope the work, sequence it and unblock it. Own proof-of-value and pilot scoping, including effort estimates and success criteria that can be measured rather than asserted. Act as incident commander for production failures. Own the root cause analysis and the remediation plan.

Advisory

Explain to finance and IT leadership what the system can and cannot do and what residual risk remains after it is deployed. Guide customers on governance, human oversight design, model risk and responsible AI.

Product and asset leadership

Convert deployment learnings into reusable assets: reference architectures, connector templates, evaluation packs and deployment accelerators. Bring evidenced feedback to the product roadmap. Distinguish a single customer’s preference from a genuine product gap.

What you bring

• 6 to 10 years building and operating production systems, including at least two years in customer‑facing delivery.
• A track record of taking at least one non‑trivial AI system to production and keeping it running afterwards.
• Strong Python and system design depth.
• Depth across LLM and agentic architectures, retrieval, evaluation methodology, guardrails and observability.
• Cloud deployment experience at scale. Azure is preferred given our stack.
• Containers and orchestration.
• Enough finance vocabulary to hold a conversation about DSO, deductions, remittance advice, and cash application without a translator or the evident ability to get there within a quarter.
• Commercial judgement. You can recognise when a customer request has quietly become unpaid scope, and you raise it early.
• Willingness to travel and embed with customers.
• Nice to have AI deployments in regulated industries.
• Enterprise ERP integration experience, particularly SAP S/4HANA or Dynamics 365 Finance and Operations.
• Experience mentoring or leading engineers.
• Pre-sales or solution architecture exposure.
• Startup or product‑building experience where customer feedback shaped what got built.

What success looks like in your first year

• Two accounts running in production against agreed KPIs.
• The Forward Deployed Engineers on your team are measurably more capable than when they joined it.
• At least two reusable assets you produced are in use by another deployment team.

If you like wild growth and working with happy, enthusiastic over‑achievers, you'll enjoy your career with us!

Experience Level Senior Level