Venkateswarlu GollaHyderabad, India

I started on the order desk. Now I design the agents that run it.

Platform and product leader in AI-led business transformation for supply chain (Source-to-Pay, Lead-to-Cash) and Finance & Accounting. Twenty years across chemicals, aerospace, FMCG, pharma and banking. Author of CVPI and CAPE.

LinkedIn Request the CAPE framework Watch the 75-second story
Human-run · todayCVPI · agents across functionsCAPE · agents across companies
The story in one order

One order. Three ways of working.

Order #4471 runs Sell to Pay and hits four everyday exceptions: a price rise, a part swap, a reroute and a wrong invoice. Same order, same exceptions, three outcomes.

Human-runPaid on day 20. 18 hours of people time spent chasing.
CVPIPaid on day 11. Agents work across functions, but stop at the company wall.
CAPEPaid on day 3. Agents work across companies, under shared rules.

Illustrative scenario with conservative numbers. Real supply chains carry far more exceptions.

The thinking

Agents act. People lead.

CVPI · Published 2025

Cross-functional Vertical Process Intelligence

Look at any business process at activity level and most of the steps repeat from function to function. Only the rules, policies and controls change. So build an AI agent once and reuse it across functions.

CAPE · Next 1–3 years, my view

Cross-enterprise Accountable Partner Ecosystem

The next wall is between companies. Buyer and supplier agents settle routine exceptions directly, under shared rules both sides agree. People make the big calls and stay accountable.

Proof points that this works at scale already exist: retailers using AI agents to negotiate with suppliers, and shared rails like UPI and FASTag that settle in seconds.

CVPI breaks the walls between functions. CAPE breaks the walls between companies.

Career, Sell to Pay

Every stage of the work, from inside it

I have run the work by hand, fixed the process, automated it, and built AI products and platforms on top of it. Each stop below taught me what the next one needed.

2025 – nowAgentic platforms

Principal Consultant, Forward Deployed AI Solutions

Infosys BPM · Hyderabad
  • Owns P&L for agentic AI and consulting in Sales and Fulfilment, with a team of 25+.
  • Designed IOOP, a composable agentic platform on Infosys Topaz with 8 front-stage and 11 background agents. It plugs into any client’s IT landscape.
  • Built the human-in-the-loop review screen: source evidence beside every field, only exceptions highlighted. Manual effort down 40–60%.
  • Took IOOP from first deployment to a reusable product. New-client go-live cut from months to about 8 weeks.
  • Designed a target operating model for agentic operations that prepares for agent-to-agent trading between companies.
2018 – 2024AI products

Senior Manager, Digital Transformation Services

Genpact · London
  • Owned the Cognitive Buying Assistant end to end: an AI front door to Source-to-Pay with guided buying, supplier self-service and a smart helpdesk.
  • Took it from 10 proofs of concept to full rollout at a global FMCG client. 3,000+ employees adopted it and buying cycle time fell by 50%.
  • Co-created ServiceNow’s Source-to-Pay capability. Led implementations on SAP Ariba, Coupa, ServiceNow and Appian.
  • Advised CFOs and process owners on spend leakage using Celonis and Signavio.
2017 – 2018Automation

Project Lead, RPA Centre of Excellence

Mphasis · Bengaluru
  • Found 40 automation opportunities and took 25 to production, cutting manual effort by 50% on those processes.
2015 – 2017Process ownership

Global Process Owner, Source-to-Pay

Xerox (now Conduent) · Bengaluru
  • Set one Source-to-Pay process model across 120 countries for a global automotive OEM, with a team of 80. Catalog spend compliance raised to 60%.
2009 – 2015Transitions

Process Lead and Team Lead, BPM Transitions

Capgemini · Bengaluru
  • Delivered 5 transitions on time and within budget, with SLAs back on target within 4 weeks of go-live.
2006 – 2009The order desk

Senior Associate, 3PL Operations

EXL Service · Bengaluru
  • Ran order fulfilment: 1,500 shipments a month at 95% on time. Halved processing time from 48 to 24 hours by fixing root causes.
Results that travel

What changed for the people doing the work

40–60%less manual effort with a human-in-the-loop review screen
~8 wksnew-client go-live on a reusable agentic platform, down from months
50%faster buying cycle, with 3,000+ employees using an AI buying assistant
120countries on one Source-to-Pay process model
25automations taken from design to production
48→24 horder processing time, fixed at the root
Toolkit

Skills and credentials

Agentic AI and product

Multi-agent orchestrationGrounding & guardrailsHuman-in-the-loop designAgentic process designSkill filesProduct roadmapBusiness case & valueGo-to-market

Transformation

Target operating modelLean Six SigmaValue stream mappingRoot cause analysisSOP & process designChange management

Platforms

Infosys TopazIBM watsonxSAP AribaCoupaServiceNowAppianSAP S/4HANACelonisSignavioSalesforceAzure AI

Education and certifications

MSc Supply Chain & Procurement Operations, Cardiff UniversityB.Com, Kakatiya UniversityPMPLean Six SigmaAgile Scrum MasterServiceNow Certified AdministratorCelonisIACCM
Request access

Get the CAPE framework slides

The full framework covers the shared rules, decision rights, earned agent autonomy and the roadmap from today’s order desk to agent-to-agent commerce. I share it one to one.

  1. Message meOn LinkedIn or by email.
  2. Write “CAPE”Add your role and company.
  3. Get the slidesI reply with the deck and am happy to talk it through.
gvenkatnaidu@outlook.com Message on LinkedIn