Hexaware Technologies Ltd. announces a new order win

The specific financial outlay for the upGrad collaboration—which focuses on training over 100 employees to become Gen AI architects ^[1]—and a dedicated "Training and Development" line item are not separately disclosed in the available annual reports or KPI data. Consequently, a direct line-item reconciliation is not possible from public disclosures, and such capability-building expenditures are subsumed within broader operating or employee benefit expenses.

Employee Benefit Expenses and Historical Trend

• Consolidated Employee Costs: Total employee benefit expenses were Rs 1,862.2 Crores in FY25 and increased to Rs 8,077.3 Crores in FY26 ^[2]. As a percentage of consolidated revenue, employee costs remained stable at 58.1% in FY25 and 58.4% in FY26 ^[3].
• Standalone Employee Costs: Standalone employee costs stood at Rs 777.60 Crores in FY25 (44.1% of standalone revenue) and Rs 3,384.1 Crores in FY26 (44.5% of standalone revenue) ^[4].

Implications and Disclosure Gaps

• Cost Classification: Specialized corporate upskilling initiatives like the upGrad partnership are typically treated as overhead or staff enablement expenses, absorbed within general employee benefit expenses or other administrative expenses without granular sub-segment reporting.
• Margin Impact: The marginal increase in employee cost intensity (up 0.3 pp on a consolidated basis YoY) reflects ongoing investments in talent and advanced technological capabilities, though individual vendor-partner outlays remain opaque ^[3].

Hexaware does not disclose a discrete baseline specific to an "AI-certified" workforce, nor does it publish quantitative milestones or KPIs specifically tied to the upGrad Enterprise partnership in its latest reported materials.

Instead, the company reports broader learning and talent development metrics under its HexaVarsity framework:

• Advanced Certifications: 44,000+ advanced certifications achieved over the prior year ^[5].
• Digital Training: 81% of Hexaware's tech workforce digitally trained ^[5].
• Learning Intensity: An average of 175.59 learning hours per employee to date ^[5].
• Program Impact: 3% lower attrition across HexaVarsity participants, with 14,000 consultants engaged through Club Synergy ^[5].

Partnership Scope and Disclosure Gaps

The collaboration with upGrad Enterprise is structured around joint upskilling, experimentation, go-to-market initiatives, and enterprise transformation targeting 'Global Programs' for AI ^[6]. However, specific numeric targets, timelines, or performance milestones that this partnership is explicitly intended to accelerate are not disclosed in available filings or press releases.

Hexaware relies on external EdTech partnerships for workforce and client AI upskilling—notably its expanded collaboration with upGrad Enterprise for enterprise-wide GenAI skilling and transformation programs ^[7]. In contrast, mid-cap IT peers predominantly utilize proprietary in-house academies supplemented by targeted academic tie-ins. Quarterly financial filings do not explicitly isolate external EdTech vendor subscription fees or internal training overhead within SG&A or operating expense lines, as training expenses are absorbed within broad employee benefit and other operating expenses.

Training Model Comparison: External EdTech vs. In-House Academies

• Hexaware: Employs an external partnership model centered on upGrad Enterprise to rapidly scale GenAI competencies across global delivery teams and client-facing transformation programs ^[7].
• Mphasis: Operates an in-house academy framework via Talent Next and the Mphasis Learning Academy, delivering structured cohort training in AI/ML and cloud, paired with internal mobility channels (Talent Navigator) ^[8].
• Zensar Technologies: Leverages internal infrastructure anchored by its proprietary ZenLearn platform and a flagship Pune Learning Academy hub, supporting early-career "Graduate to Professional" tracks and LeAP leadership programs ^[9].
• KPIT Technologies: Focuses on specialized, domain-specific academic partnerships, such as its workplace Master of Technology program in automotive software engineering co-developed with Coventry University ^[10].

Observed Impact on Margins and Cost Structures

Quarterly financial disclosures do not disaggregate training delivery costs from broader operational overhead, making a direct line-item quantification of external EdTech versus in-house training cost impact unfeasible. However, a review of recent quarterly KPI data across the peer group shows that overhead and employee cost ratios remain governed by utilization and wage inflation rather than training delivery architecture:

• Hexaware (Consolidated): Employee costs ranged from 56.9% to 59.4% of revenue across Q1 FY26 to Q4 FY26 ^[11], while consolidated EBITDA margins tracked between 16.4% and 17.6% ^[12]. Other operating expenses ranged from Rs 896.50 Crores to Rs 1,060.00 Crores ^[13].
• Mphasis (Consolidated): Employee cost percentages ran between 54.4% and 57.5% ^[14], with consolidated EBITDA margins sustaining between 20.3% and 21.5% ^[15].
• Zensar Technologies (Consolidated): Employee costs accounted for 62.6% to 65.1% of revenue ^[16], with consolidated EBITDA margins moving between 18.9% and 21.9% ^[17].

Analytical Implications

The choice between external EdTech outsourcing (Hexaware) and proprietary academies (Mphasis, Zensar) reflects divergent go-to-market execution rather than a measurable SG&A margin divergence. External partnerships offer rapid scalability for broad-based GenAI adoption without upfront physical or curriculum-development overhead, whereas in-house academies secure deep proprietary domain alignment at the expense of fixed internal L&D maintenance overhead. Because neither model isolates training costs as a distinct financial line item in statutory filings, any efficiency gains or margin drag from platform licensing fees versus internal faculty costs remain obscured within general SG&A and employee benefit lines.

_Scope note: this comparison also included Tata Elxsi Ltd. (TATAELXSI); Fractal Analyt. (FRACTAL), which the answer above does not cover. Ask about any of them for a full side-by-side._