Freshworks 2026: What AI-First Engineering Means for Chennai Freshers

In May 2026, Freshworks CEO Dennis Woodside said ‘over half of our code is originated in AI today,’ then announced the company’s third round of layoffs in 30 months.

That’s not a warning. It’s the clearest available data point on what AI-first engineering looks like inside a Chennai-born product company, and what skills sit on the right side of that shift.

Two Rounds, One Data Point

In November 2024, CEO Dennis Woodside wrote to employees announcing a reduction of 660 employees globally: around 500 in India (12 percent of the India workforce) and 16 percent of the US workforce. The stated reason was a strategic realignment around employee experience, AI, and customer experience priorities.

Eighteen months later, Freshworks reported revenue growing 16 percent year-on-year in Q1 2026 while simultaneously announcing around 500 more cuts from an end-2025 headcount of approximately 4,500 full-time employees. Woodside’s words from the Q1 2026 earnings call:

• ‘Over half of our code is originated in AI today… that is definitely changing the amount of people we need to build product.’
• ‘We will be very thoughtful about any additional headcount or backfilling existing roles.’

Growing revenue, shrinking headcount. That’s the AI productivity story in operational terms, not conference-keynote terms. The question for a 2026 engineering fresher is which engineers end up on the ‘doing more’ side of that equation.

What Freddy AI Actually Does

At the June 2025 Refresh event, Freshworks launched the Freddy Agentic AI Platform: AI agents that resolve service requests end-to-end, not just route or answer them. The platform includes:

• Freddy AI Agent Studio (no-code builder for custom agents)
• Freddy AI Agent for Email (autonomous email resolution)
• Freddy AI Copilot with Reply Suggester (assisted human response)
• Stated performance: Freddy AI Agent resolves 45 percent of CX requests and 40 percent of EX (IT helpdesk) requests autonomously on average

In November 2025, Freshworks added Freshdesk Command Center (a unified multi-channel workspace with AI assistance), Vertical AI Agents for eCommerce, fintech, travel, and logistics, and Freddy AI Insights for real-time service trend visibility.

The Vertical AI Agents are a useful signal about where the engineering work is now concentrated: building domain-specific agent behaviour, not general-purpose chat functionality. An engineer who understands how to define a tool call, design an agent fallback, or test edge cases in an agentic workflow is working in the territory Freshworks’s own product teams are actively building in.

The engineering work that remains after the agentic layer handles routine requests:

• Building agent orchestration and tool call integration
• Evaluating agent outputs for correctness before merging to production
• Extending the platform for vertical and enterprise use cases

That’s the work AI has not automated away.

The Verified 2025 Fresher Tracks

Freshworks ran two Graduate Trainee tracks out of Chennai for the 2025 batch. Here’s what the official listings confirmed.

Graduate Trainee: Customer Support

Item	Detail
Location	Chennai (5 days per week, in-office)
Shifts	Rotational (mandatory)
Eligibility	2025 batch only, no standing arrears
Screening	Assignment submitted within 1 week of shortlisting
Selection events	2 in-person Saturday events at Chennai office
CTC	Not officially disclosed by Freshworks

The CTC for this role is not published by Freshworks. The SmartRecruiters listing confirms the selection process and eligibility but states no figure. Aggregator sites estimate 5 to 6 LPA for the 2025 cycle; treat that as a directional ballpark, not a floor or ceiling.

The role is customer-facing, using Freddy AI tools day-to-day. This is the support track that works alongside the product, not the engineering track that builds it.

Graduate Software Trainee

The engineering track ran a coding challenge followed by two to three rounds of virtual Zoom interviews across multiple drive dates. Eligibility: BE/BTech/BCA, no active backlog, strong DSA and object-oriented programming. The squad model means the engineer owns full SDLC: backend, frontend, testing, and operations.

CTC for this role is not verifiable from any primary source as of May 17, 2026. Do not accept aggregator figures for the engineering track without a primary-source citation.

2026 Status

No 2026 batch fresher cycle has been announced as of May 17, 2026, twelve days after the May restructuring. Woodside’s explicit caution about new headcount makes near-term announcements unlikely. Monitor the official Freshworks careers portal and LinkedIn page rather than aggregator listings, which often carry stale or unverified information.

Positioning for the AI-Era Product Company Track

Placement preparation for a product company like Freshworks differs from what gets a fresher through TCS Ninja or Infosys InfyTQ.

Service company assessments (AMCAT, cocubes, NQT-style tests) measure breadth: aptitude, basic coding, verbal ability. Product company selection measures depth: DSA fluency, full-stack problem decomposition, and demonstrated ability to work alongside AI tooling.

What ‘working alongside AI tooling’ means in practice at a company where more than half the code is AI-originated:

• Reviewing AI-generated code for edge cases and correctness before merging to production
• Writing prompts or tool configurations that produce correct output for a specific task
• Integrating LLM API responses into a feature rather than hand-coding the feature entirely
• Debugging when an AI agent misroutes a ticket or returns a wrong answer

Most engineering programmes train students to write code from scratch. That skill is still necessary. It is no longer sufficient at a company that has already automated a meaningful share of routine coding tasks.

The practical implication: strong DSA and OOPs remain the technical entry requirement for the engineering track. Product companies still screen for algorithmic thinking first. The layer that separates two candidates with similar fundamentals is demonstrated evidence of working with AI tooling. A project on GitHub that integrates an LLM API answers a question that a list of certificates cannot: has this person actually shipped something with AI, or only read about it?

If aptitude preparation for service company assessments is already in progress, that work carries over. DSA and OOPs form the base for both tracks. The AI skills portfolio is the differentiator for product company shortlisting specifically. Building both in parallel is the practical approach for a 2026 final-year student who is keeping product company options open alongside service company targets.

For the specific tools and projects that make an AI-skills portfolio credible in 2026 product company interviews, FACE Prep’s AI roadmap for Indian engineering students lays out the curriculum by role and placement timeline. For background on Freshworks’s product suite and hiring history, the Freshworks company and product overview covers the full picture.

Woodside’s ‘over half our code is AI’ statement is not a figure of speech. It is a concrete signal about what the company values in the engineers it still hires. The ones who demonstrate that they can direct a coding agent, catch what it misses, and ship something that holds up in production are on the right side of that statement.

TinkerLLM is where you build that working-with-AI literacy without the setup overhead. ₹299 puts real LLM API calls in your hands, and the micro-project that results is what goes on the GitHub profile a product company recruiter checks during shortlisting.