Genpact

By FY2018, Genpact operated large-scale back-office processes for Fortune 500 clients — accounts payable, order management, claims processing, and financial closing. These operations were labor-intensive: approximately 70% of delivery cost was direct labor (processors, analysts, quality checkers). Rising wages in India (7-10% annual wage inflation for experienced BPO staff) were compressing margins. The company's Lean Six Sigma heritage had already optimized manual processes extensively, leaving limited further gains from process redesign alone. Gross margin was approximately 34.5%, and the company needed a structural cost lever beyond headcount management.

Genpact deployed its "Lean Digital" framework, combining process expertise with automation technology:

• Process mining and task decomposition: Before deploying automation, Genpact used its proprietary Cora platform to map and decompose client processes at the task level. Each process was broken into discrete steps, classified as "automatable" (rules-based, repetitive, structured data) or "human-required" (judgment, exception handling, unstructured data). This task-level mapping covered over 200 distinct process types.
• RPA at scale: Deployed over 5,000 software bots across client operations by FY2021, automating tasks such as invoice data extraction, three-way matching, journal entry posting, and claims adjudication. Each bot replaced approximately 2-3 FTE equivalents of manual processing.
• AI-assisted decision-making: Layered machine learning models on top of RPA for processes requiring judgment. For example, in accounts payable, AI models predicted exception types and suggested resolutions, reducing human review time by approximately 40% on exception items. In claims processing, NLP models extracted data from unstructured documents, eliminating manual keying for ~60% of incoming claims.
• Continuous improvement loop: Built real-time dashboards tracking automation coverage, bot utilization, and error rates per process. Monthly reviews identified new automation candidates based on exception patterns and volume trends.

• FTE equivalent savings: Automation eliminated approximately 12,000-15,000 FTE equivalents of manual delivery work across the company's operations by FY2021 — roughly 10-13% of the delivery workforce.
• Cost-per-transaction reduction: Average cost per transaction declined approximately 25-30% in fully automated processes (e.g., invoice processing dropped from approximately $3.50 to $2.40 per invoice).
• Gross margin improvement: Gross margin expanded from 34.5% in FY2018 to 36.2% in FY2021, a 170 basis point gain, with automation cited as the primary driver on earnings calls.
• Quality improvement: Error rates in automated processes dropped by approximately 50%, from ~2.5% to ~1.2%, as bots eliminated manual keying errors.
• Client SLA improvement: On-time delivery rates improved from approximately 94% to 98% across automated processes.
• Timeframe: FY2018-FY2021 (3-year scale-up).

• Lean Six Sigma heritage meant processes were already standardized and documented — essential prerequisites for automation
• Proprietary Cora platform provided task-level process intelligence that off-the-shelf RPA tools couldn't replicate
• GE-legacy operational rigor enabled disciplined bot deployment with proper testing and monitoring
• Client willingness to share in automation savings through gain-sharing contracts incentivized investment