What technology stack does Wells Fargo run, and how is AI built on it?
Wells Fargo runs a hybrid stack it chose in September 2021: Microsoft Azure as primary public cloud, Google Cloud for 'advanced workloads, and complex artificial intelligence and data solutions', and a move into third-party-owned data centers offering private cloud and traditional hosting, with a long-term aim to 'rely predominantly on public cloud'. The 2026 proxy records 'completion of new data centers', and the 2026 investment plan is to keep migrating into them and moving applications to public and private cloud, while modernizing core banking and call-center platforms with AI. AI runs on Tachyon, an internal platform CIO Chintan Mehta said rests on three assumptions: no single model will win, no single cloud, and data will be hard to move. Fargo shows the pattern: Google Dialogflow, Gemini Flash 2.0 for intent only, a small in-house model to strip personal data, and an orchestration layer that keeps APIs off the model path; Llama and OpenAI models serve other uses. Employees get Google Agentspace and Microsoft 365 Copilot. Technology, telecommunications and equipment expense was $5,203 million in 2025, up 14%, driven by amortization of internally developed software, software licenses and hardware. Head of technology Bridget Engle runs a budget of more than $10 billion and 30,000 technologists.[1][2][3][4][5][6][7][8][9][10][11]
Data as of Oct 31, 2026 · Wells Fargo profile updated Oct 21, 2026 · 25 sources cited on this page · 141 banks profiled on this site · Markdown
Which clouds, platforms and models make up Wells Fargo's stack?
What Wells Fargo or a named executive has said about each layer. 'Since' is the first dated mention.
| Layer | What Wells Fargo uses | Since | Who said it |
|---|---|---|---|
| Public cloud | Microsoft Azure as primary provider for strategic business workloads; Google Cloud for advanced workloads and complex AI and data solutions | September 15, 2021 | Saul Van Beurden, then head of technology[1] |
| Data centers and private cloud | Transition to third-party-owned data centers with private cloud and traditional hosting; 'completion of new data centers' in 2025; 2026 plan to keep migrating and move applications to public and private cloud | 2021; 2025-2026 | Wells Fargo release; 2026 proxy; 4Q25 presentation[1][2][3] |
| AI platform | Tachyon: built for many models, more than one cloud and data that is hard to move; supports model and tensor sharding; described in payments as 'model and cloud agnostic' | January 2024 | Chintan Mehta, CIO; Ather Williams, payments[4][12] |
| Customer assistant | Fargo on Google Dialogflow; launched on PaLM 2, later Gemini Flash 2.0 for intent and entities only; local speech-to-text and a small PII-detection model before any model call | October 2022; April 2025 | Wells Fargo; Mehta[5][6][4][7] |
| Other models | Meta's Llama for some internal uses; OpenAI models 'can be tapped as needed'; orchestration picks the model per task ('compound systems') | 2024; 2025 | Mehta[4][7] |
| Employee tools | Google Agentspace, Gemini Deep Research and NotebookLM for all employees; Microsoft 365 Copilot for more than 30,000 staff with 92% active use; an in-house developer platform for use cases | June-November 2025 | Tracy Kerrins; Ken Meyer[8][9][13] |
| Data | A centralized data team building a cloud-based data layer with authoritative sources; 2026 plan to 'invest in data platforms to better leverage AI' | November 2025; January 2026 | Van Beurden; 4Q25 presentation[14][3] |
| Specialized models | The model risk team built a small language model that beats off-the-shelf models on debit card fraud detection | October 20, 2026 | Faraz Shafiq[15] |
| Commercial digital platform | Vantage: about $2 trillion a month for roughly 92,000 companies and 1.7 million users; micro-front ends personalize by user role; embeds in Oracle, NetSuite, Microsoft, SAP and Workday; traffic up 222% in 2025 | Launched 2021 | Reetika Grewal, Wholesale Digital[16] |
How much does Wells Fargo spend on technology, and how many technologists does it have?
Expense lines are from the 10-K income statement; the budget and headcount are as Bridget Engle described them to American Banker.
| Measure | Value | As of |
|---|---|---|
| Technology, telecommunications and equipment expense | $5,203 million, up 14% from $4,583 million in 2024 ($3,920 million in 2023) | Full year 2025[10] |
| Why it rose | Higher amortization of internally developed software, higher software maintenance and license expense, higher hardware depreciation | Full year 2025[10] |
| Technology budget and team | More than $10 billion a year; a global technology team of 30,000 | October 1, 2026[11] |
| Wealth technology | More than $1 billion over several years to modernize the adviser platform, including Advisor Gateway with generative AI (launched Q2 2026) | July 14, 2026[17] |
| Software engineering productivity | Scharf: AI made engineers 30% to 35% more efficient; Engle: on top of software-engineering productivity already baked in last year, 'we've seen that like more than double' | December 2025; October 2026[18][11] |
Does Wells Fargo build or buy its AI technology?
Both, by layer. It buys cloud capacity and foundation models from Microsoft and Google and uses BlackRock's Aladdin Wealth technology for adviser tools, but builds the orchestration, privacy filters and platform around them: Mehta said 'all the computations and detokenization, everything is on our end'.[1][19][7]
The platform is designed to swap models. Mehta's 2024 description of Tachyon assumes 'one AI model won't rule the world', and Shafiq's October 2026 hub-and-spoke model centralizes governance, security and controls while business teams own their applications.[4][15]
Not disclosed: the share of applications in public cloud, the data-center operators, the core banking vendors, model sizes, GPU counts or AI spend as a separate line.[10][3]
What does the Wells Fargo profile record on this topic?
The rows below are the entries of the Wells Fargo AI strategy profile that bear on this page, as of Oct 21, 2026; the profile holds the full record of 44 dated moves, 17 use cases and 60 sources.
| Use case | What the record shows | Status | Source |
|---|---|---|---|
| Agentspace agents for employees | Agents to summarise complex FX inquiries and navigate compliance policies; research tools for all staff. | Rolling out | [20][21] |
| AI coding assistants | 30–35% efficiency gain for engineers. | In production | [18] |
| Microsoft 365 Copilot | More than 30,000 employees since June 2025; 92% active usage among enabled staff, used for content, summaries and research. | In production | [9] |
| Vendor-contract agent | A custom Agentspace agent to query roughly a quarter of a million vendor-agreement documents for clauses, payment terms and contract types. | Rolling out | [8][13] |
| Date | Move | Detail | Source |
|---|---|---|---|
| Oct 1, 2026 | Engle: $10 billion technology budget, nearly 200,000 staff with AI tools | The technology team of 30,000 is expanding cloud-based data-centre capacity for AI work; Engle holds two pending generative-AI patent applications filed in 2026. | [11] |
| Sep 29, 2026 | Cyber chief on AI-enabled social engineering | Sarah Gosler, head of Cyber Human Defense, says AI lowers the barrier to cybercrime and 'humans now represent the largest attack surface'; Wells Fargo advises clients to equip staff with AI defences. | [22] |
| Jul 15, 2026 | AI Teammate for wealth advisers | Plain-language chat over the Advisor Gateway platform; users created thousands of prompts soon after launch, per chief product officer Evelyn Varner. | [23] |
| Jul 14, 2026 | Q2 2026 call: Advisor Gateway with generative AI; headcount 197,000 | More than $1 billion spent modernising the wealth platform; headcount down 79,000 in six years; CFO sees more automation ahead. | [17] |
| May 27, 2026 | Andre Mansour hired as head of AI for wealth | Joins from Google, where he led AI work for capital markets and asset management clients; reports to Evelyn Varner. | [24] |
| May 5, 2026 | Payments head: 335 AI experiments, 26 in production | Ather Williams says the AI platform enables 85% of use cases and a call-summary tool supports bankers; 'Everyone has Co-pilot at their desk. Everyone messes around with Claude.' | [12] |
| May 5, 2026 | Payments AI platform is model- and cloud-agnostic | Ather Williams at American Banker's Payments Forum. | [12] |
| Dec 9, 2025 | AI makes engineers 30–35% more efficient | Scharf at the Goldman Sachs conference; more job cuts and higher severance expected. | [18][25] |
| Nov 3, 2025 | Kerrins at Money20/20: developer platform plus Google employee tools | Kerrins says the bank runs a developer platform and a Google partnership for employee capability; Deep Research and NotebookLM reached about 2,000 employees first. She flags agent-to-agent banking and customer 'AI buddies' as next. | [13] |
| Aug 5, 2025 | Agentic AI to all 215,000 employees | Google Cloud Agentspace, Gemini Deep Research and NotebookLM; Kerrins: 'a big deal for us'. | [20] |
| Jun 1, 2025 | Microsoft 365 Copilot rolled out to employees | More than 30,000 users and a 92% active usage rate among enabled employees by November 2025, per CIO for Enterprise Functions Ken Meyer. | [9] |
| Jan 1, 2022 | Fargo to be powered by Google Cloud AI | The virtual assistant is built on Google's conversational product. | [20][21] |
Which cloud does Wells Fargo use?
Both Microsoft Azure and Google Cloud, chosen in September 2021: Azure is the primary public cloud for strategic business workloads and Google Cloud handles advanced workloads and complex AI and data solutions. The bank also moved into third-party-owned data centers offering private cloud and traditional hosting, recorded the 'completion of new data centers' in its 2026 proxy, and plans to keep moving applications to public and private cloud.
What is Wells Fargo's Tachyon platform?
Tachyon is the internal platform Wells Fargo built to run its AI applications. CIO Chintan Mehta said in January 2024 that it rests on three assumptions: that no single AI model will dominate, that the bank will not run on a single cloud, and that data will be hard to move between stores. It supports model and tensor sharding, and the bank's payments chief describes the AI platform as model and cloud agnostic.
How much does Wells Fargo spend on technology?
Its 10-K reports technology, telecommunications and equipment expense of $5,203 million for 2025, up 14% from $4,583 million in 2024, driven by amortization of internally developed software, software licenses and hardware depreciation. Head of technology Bridget Engle told American Banker in October 2026 that she runs a budget of more than $10 billion a year and a 30,000-person technology team; the bank does not break out AI spending.
Which AI models does Wells Fargo use?
Several, chosen per task. Fargo uses Google's Gemini Flash 2.0; Meta's Llama serves some internal uses; OpenAI models can be called as needed; employees use Google Agentspace, Gemini Deep Research, NotebookLM and Microsoft 365 Copilot. The model risk team has also built its own small language model for debit card fraud detection.
- Wells Fargo Announces New Digital Infrastructure Strategy and Strategic Partnerships With Microsoft, Google Cloud
- Wells Fargo & Company 2026 Proxy Statement (DEF 14A)
- 4Q25 Financial Results presentation: areas of focus for 2026 investments (Form 8-K Exhibit 99.3)
- Wells Fargo's assistant, powered by Google's AI, poised to hit 100 million interactions annually
- Wells Fargo's New Virtual Assistant, Fargo, to Be Powered by Google Cloud AI
- Wells Fargo's New Virtual Assistant, Fargo, to Be Powered by Google Cloud AI
- Wells Fargo's AI assistant just crossed 245 million interactions – no human handoffs, no sensitive data exposed
- How Wells Fargo is using Google Cloud AI to empower its workforce with agentic tools
- From legacy to Frontier: How 100-year brands are leading AI innovation
- Wells Fargo & Company 2025 Annual Report (Form 10-K, Exhibit 13): risk factors
- The Most Powerful Women in Banking, #11, Bridget Engle, Wells Fargo
- How Wells Fargo deploys AI in payments
- How Wells Fargo is building agentic AI into its workflows
- Saul Van Beurden On How Wells Fargo Scaled Trust, AI And Growth
- 4 keys to Wells Fargo's strategy to transform the bank with AI
- Wells Fargo's Digital Platform Aims to Help Clients Do Less, Not More (Vantage, Innovation of the Year)
- Wells Fargo (WFC) Q2 2026 Earnings Call Transcript
- Wells Fargo CEO Sees AI Impacting Companies' Workforce Decisions
- Wells Fargo (WFC) Q3 2026 earnings call transcript
- Wells Fargo plans to unleash Google's agentic AI bank-wide
- Wells Fargo hires AWS exec to lead AI strategy
- Why AI is a Corporate Treasurer's Best Friend
- Wells Fargo debuts AI-powered teammate for financial advisers
- People Moves: Wells Fargo Hires Former Google AI Finance Leader
- Wells Fargo expects more job cuts, will roll out AI gradually in 2026
When Wells Fargo moves on AI, you'll read it here first.
when one of the banks we cover moves on AI, it is in the next morning's brief · six sourced stories · 7 am PT · free
plus every regulation, bank and agent page update, the morning after · leave any morning