Enterprises are moving quickly toward AI adoption, often without a unified control strategy. Nobi Advisory Services helps leadership teams define a secure, scalable path forward before architecture decisions are locked in. We work with CIOs, CISOs, and enterprise architects to assess readiness, identify exposure, and design a practical AI governance roadmap aligned to business goals.

• AI readiness and risk assessment
• Identity and access architecture review
• Agent governance strategy design
• Control framework definition
• Executive and board-level AI risk briefing
• Roadmap definition and prioritization

You design how people move through the NobiTrust platform. Enterprise security consoles are notoriously dense — policies, audit trails, agent telemetry, entitlement graphs — and most of them leave people hunting for the one screen they actually needed. Your job is to make browsing the platform obvious: map the journeys, cut the steps out of the paths people repeat every day, and hand our Full Stack Developer wireframes precise enough to build from without guesswork. The role reaches past the product, too — you own the catalog and the customer deck that explain NobiTrust to the people evaluating it.

What you will do

• Map the console's current journeys and find where users get lost
• Simplify navigation and information architecture across the platform
• Produce wireframes and interaction specs engineering builds directly from
• Own the design system's visual language with the engineer implementing it
• Design the product catalog shared with prospective customers
• Build and maintain the customer-facing deck used in demos and briefings
• Test flows with real users and fold what you learn back into the roadmap
• Hold the product to WCAG 2.2 AA and keyboard-first operation

What we are looking for

• A portfolio of shipped product UX, not only visual concepts
• Wireframes and specs an engineer can build from unaided
• Fluency in Figma, including components, variants and prototypes
• A track record simplifying dense, information-heavy enterprise tools
• Real command of layout and typography, on screen and in a deck
• You work alongside engineers rather than handing files over a wall
• Comfortable producing customer-facing collateral, not just product screens
• Enterprise security, identity or developer-tool experience is a plus

You own the model layer that decides whether an agent's next action is safe. That means intent classification on tool calls, prompt-injection and exfiltration detection, and risk scoring — all running inline, inside a latency budget measured in milliseconds, not seconds. This is applied LLM engineering under production constraints: evaluation harnesses matter more than demos, and a model that is right 99% of the time but unbounded in tail latency is not shippable.

What you will do

• Design and ship the inline guardrail and intent-classification pipeline
• Build red-team and regression eval harnesses for adversarial prompts
• Fine-tune and distill small models to hit inline latency targets
• Build retrieval over policy, schema, and entitlement corpora
• Instrument model decisions so every verdict is auditable and explainable
• Partner with the gateway team on inference at the enforcement path

What we are looking for

• Strong Python and comfort in a production codebase, not just notebooks
• Hands-on transformer fine-tuning — LoRA / PEFT, distillation, quantization
• Rigorous about evals: you build the harness before you trust the number
• Serving experience with vLLM, TensorRT-LLM, or equivalent
• Fluent with frontier model APIs and tool-use / MCP integration patterns
• Security-adjacent ML or adversarial robustness work is a strong plus

You own the console — the interface where a security architect writes the policy that governs every agent in their enterprise, then watches it take effect in real time. This is deliberately a UI-weighted role: roughly four days in five are front-end craft, and the remaining slice is the API work needed to serve those screens properly. Dense, high-stakes, information-heavy interfaces that people keep open all day, built to a standard where the details are obvious in use and invisible in description.

What you will do

• Build the policy authoring experience — readable, diffable, hard to misuse
• Ship live telemetry and audit-trail views over high-volume agent traffic
• Own the design system, component library, and interaction patterns
• Hold the product to WCAG 2.2 AA and keyboard-first operation
• Write the Java and Postgres endpoints behind your own screens
• Take features end to end — design conversation through production

What we are looking for

• Expert React and TypeScript, plus modern CSS without a framework crutch
• A track record of data-dense dashboards, tables, and time-series views
• Real visual taste — you notice spacing, rhythm, and hierarchy unprompted
• Comfortable writing backend code in Java or Node against PostgreSQL
• Performance instincts: virtualization, streaming updates, render budgets
• Security or infrastructure product experience is a plus

You are the technical voice in the room. Enterprise security teams will not buy an enforcement layer on a pitch deck — they need someone who can whiteboard the architecture against their actual identity provider, their actual gateway, and their actual agent framework, then prove it in a proof of concept. You run discovery, own the technical win, and carry what you learn in the field straight back into the roadmap.

What you will do

• Run technical discovery with security, identity, and platform teams
• Deliver demos tailored to the customer's stack, not a canned script
• Scope, design, and drive proofs of concept to a defined success test
• Own security questionnaires, RFP responses, and architecture reviews
• Advise on integration across IdPs, API gateways, and agent frameworks
• Convert field signal into prioritized, evidence-backed product input

What we are looking for

• 3+ years as a solution / sales engineer in security or infrastructure
• Genuine fluency in OAuth 2.0, OIDC, SAML, and token architecture
• Working knowledge of API gateways and Zero Trust access models
• Enough hands-on ability to script and stand up a PoC yourself
• Clear, credible communication with engineers and with a CISO alike
• Early-stage instincts — you build the playbook rather than inherit one