*Forward Deployed Engineer Archetype:** Technical lead · Trusted advisor · Practice builder Role Summary
You are the CTO\-equivalent inside a client engagement. You set AI strategy, architect enterprise\-grade agentic systems, and create patterns that scale across Cognizant's entire client portfolio. Your influence extends far beyond the account you are sitting in — you are simultaneously solving for this client and building the practice. What You Will Do
Act as the primary technical authority on agentic AI engagements — owning architecture, delivery quality, and client executive relationships simultaneously across one or more strategic accounts.
Define the client's agentic AI roadmap: identify high\-value use cases, sequence the build, set the success metrics, and tie everything back to business outcomes (cost, revenue, risk, speed).
Architect enterprise\-scale multi\-agent systems — including agent orchestration topology, inter\-agent communication protocols, memory and state management at scale, fail\-safe and escalation paths, and governance / audit layers.
Drive build\-vs\-buy decisions, model selection, and platform choices with a clear\-eyed view of client constraints (security, compliance, data sovereignty, existing tech stack).
Establish evaluation and observability standards for the engagement: evals infrastructure, drift detection, model performance monitoring, agent behaviour auditing, and rollback strategies.
Represent Cognizant at CTO/CAIO\-level discussions; translate complex AI system behaviour into board\-ready risk and value narratives.
Identify repeatable deployment patterns across engagements and productise them as accelerators, reference architectures, and reusable agents within the AI Market Unit.
Recruit, mentor, and develop a high\-performance FDE team; raise the capability bar of the entire practice through internal tech talks, architecture reviews, and hands\-on coaching.
Contribute to Cognizant's external thought leadership — conference talks, technical blogs, industry working groups — as a recognised voice in enterprise agentic AI. Technical Mastery Required
Deep expertise in at least two agentic frameworks; has extended or contributed to open\-source AI tooling
Multi\-cloud fluency (AWS Bedrock, Azure AI Foundry, GCP Vertex AI); knows where each excels and fails
Fine\-tuning, RLHF, and model adaptation strategies for enterprise domains
Distributed systems fundamentals: how agent systems behave under load, partial failure, and adversarial inputs
AI security and red\-teaming: prompt injection, jailbreaks, data leakage vectors, and mitigations Leadership \& Impact Requirements
Track record of multiple successful GenAI / agentic AI production deployments at enterprise scale
Has built or scaled an engineering team (hiring, mentoring, technical career development)
Executive presence: credible in the room with CTO, CDO, and business unit heads
Has shaped product strategy — turned field learnings into platform features or accelerator IP
Demonstrated ability to navigate organisational resistance and drive AI adoption across large enterprises What Makes You Stand Out
You have gone from 'AI won't work in our environment' to production rollout at a Fortune 500 — and you can tell the story concisely
You build things that outlast your engagement: not just working software, but the client team's capability to maintain, extend, and govern it
You are the person other FDEs call when something breaks at 2 AM and nobody knows why Craft Standards — All FDE Levels
These behaviours are non\-negotiable across all three levels. They define the culture of the Cognizant AI Market Unit FDE practice and differentiate our engineers from traditional consultants or staff augmentation engineers. Builder Identity
You write, debug, and ship production\-grade code. FDEs are not slide\-deck architects; you are hands\-on\-keyboard engineers whose credibility comes from working software.
You prototype in hours, not weeks. Speed of iteration is a core skill, not a nice\-to\-have.
You leave codebases cleaner than you found them. You write tests, document decisions, and build for the next engineer, not just the demo. Client Orientation
You ask 'what does success look like in 90 days?' before you write a line of code. Business outcomes drive every technical choice.
You communicate uncertainty honestly and early. Clients trust you more when you say 'I don't know yet, here's how I'll find out' than when you pretend to have all the answers.
You adapt to the client's environment — their security posture, tech stack, and organisational culture — rather than imposing a one\-size\-fits\-all solution. Agentic AI Standards
You treat evaluation as engineering. Every agent you deploy has defined evals, acceptance criteria, and an observability story before it goes to production.
You build with safety and governance in mind from day one: guardrails, audit trails, human escalation paths, and data privacy controls are not optional add\-ons.
You are framework\-agnostic and model\-agnostic. You choose the right tool for the job, not the tool you know best. Practice Citizenship
Every engagement generates at least one reusable asset — an accelerator, a reference architecture, a prompt library, or a lessons\-learned document — contributed back to the AI Market Unit.
You leave clients capable of continuing without you. Knowledge transfer and team enablement are success criteria on every engagement.