Forward Deployed Engineer at Cognizant

Cognizant hires Forward Deployed Engineers in Bengaluru and Chennai to build and ship AI agents and generative AI applications on-site with clients. The roles were listed on Cognizant’s careers site at four levels, from associate to senior architect. Several of those pages now show the title “AI Customer Engineer”, so the FDE job at Cognizant may appear under either name.

What the FDE does at Cognizant

The engineer-level listings (job 00067697575, published 3 June 2026, and job 00069053831, published 19 May 2026) describe the same hands-on job. The FDE:

• Builds and deploys AI agents and GenAI apps for clients, using agent frameworks such as Google ADK, Copilot Studio, CrewAI or AutoGen.
• Designs RAG and LLM integrations, including prompts, tools, guardrails and vector databases.
• Writes the backend. Scalable services and REST APIs, typically with FastAPI, in a microservices setup.
• Ships to the cloud with containers and cloud-native deployments, and keeps the result reliable and secure.
• Sits with the client. The listings call it an on-site role, five days a week, in a client or Cognizant office in Chennai or Bengaluru.

The senior architect role (job 00067697901, published 30 March 2026) sits above this. It sets the generative and agentic AI architecture for an enterprise, designs multi-agent systems end to end, leads proofs of concept and MVPs, and owns AI governance and standards.

Why Cognizant is building this team

Cognizant has tied the FDE model to its cloud partnerships. In an April 2026 Mission Media report on its Google Cloud partnership, Cognizant’s Annadurai Elango described “forward-deployed engineers working directly alongside Google Cloud’s teams”.

Two larger 2026 commitments show where the work is heading:

• In July 2026 Cognizant committed to scale to 5,000 “Frontier-certified” engineers and 10,000 business operators, with the first cohort due by Q4 2026, according to NewsBytes.
• In October 2026 it formed an OpenAI practice of more than 25,000 people, mostly existing staff realigned rather than new hires, according to FourWeekMBA. The same piece says Cognizant ran Codex hackathons in India and the Americas with more than 12,000 participants.

Neither report gives an FDE headcount or says how many of these roles will be in India. Read them as a sign of demand, not a hiring number.

Experience and skills Cognizant asks for

Three things stand out:

• Skills, not years, at engineer level. The India engineer listings name no minimum years. They call GitHub a must and want working agents, RAG and APIs. A portfolio of shipped projects matters more than tenure. Our FDE skills guide covers how to build one.
• It is fully on-site. Five days a week at a client or Cognizant office. If you want remote work, this isn’t it.
• Architects need enterprise platform knowledge. The senior listing prefers experience with Salesforce Einstein GPT, ServiceNow Now Assist, SAP Joule, Workday AI or Oracle AI Services.

An associate-level FDE role in Bengaluru and Chennai (job 00067697574) also appeared in search results, but we couldn’t open the page to confirm its requirements.

Pay

Cognizant doesn’t post pay on its India FDE roles. Its US AI Forward Deployed Engineer listing, seen on Match Moth in September 2026, posts US$130,000–158,500 a year. That is a US figure and doesn’t indicate India pay.

As an India estimate, here is Cognizant’s general software engineer pay from Levels.fyi, updated 7 November 2026:

For senior architects, AmbitionBox shows a range of ₹15.2–45.3 LPA for 15–22 years of experience, from 487 salaries, but that page was last updated in December 2024. Cognizant doesn’t say which level its FDE roles map to, so treat all of these as estimates. Compare bands across employer types in our FDE Salary Estimator.

How to prepare

1. Build one working agent with a real framework. Pick Google ADK, CrewAI or AutoGen, connect it to two tools and a vector database, and add guardrails. These frameworks are named in the listing.
2. Wrap it in a FastAPI service and a Docker container. Push it to GitHub with a clean README, tests and a short demo. The listing calls GitHub a must.
3. Learn one cloud AI platform well. Cognizant works with Google Cloud, Microsoft, AWS and OpenAI. Deploy your project on one of Vertex AI, Azure AI Foundry or Bedrock and note the cost and latency.
4. Practise a client conversation. Some Cognizant roles include client interviews. Rehearse turning a vague business problem into a scoped AI solution, with risks and a timeline.
5. Aim at architect level only with the years to match. The architect role wants 12+ years and TOGAF-style enterprise architecture. If you are earlier in your career, target the engineer roles first.

Compare Cognizant with other IT services FDE programmes on our Accenture, Infosys and HCLTech pages, or see openings in Chennai.

Cognizant FDE interview: what is publicly known

1. Application: India FDE roles are on Cognizant's careers site under 'Forward Deployed Engineer' links, but some pages now display the title 'AI Customer Engineer' or 'GenAI - Senior Architect'. Search for both.
2. Recruiter call: Cognizant's published hiring process starts with an introductory call with a recruiter after your application is reviewed.
3. Hiring team interview: Cognizant says some roles also need technical assessments and/or client interviews. For these client-facing roles, expect hands-on questions on agents, RAG, APIs and Docker. Interviews may be in person or by video.
4. Decision: Cognizant's process ends with a hiring decision and offer. It doesn't publish an FDE-specific format, so ask the recruiter how many technical and client rounds to expect.