# Nova Intelligence

Aiolum story: https://aiolum.com/story/nova-intelligence/ · added 2026-11-06 · updated 2026-11-06

**In short.** Nova Intelligence (AI agents that document, change, test and run the SAP systems of large companies, United States): Festo and KION named as SAP customers, May 2026. Evidence grade E: no revenue figure was checked. Team: 3 co-founders (Emma Qian, Sam Yang, Prof. Dr. Alexander Zeier), founded 2024; Fortune reported "still just 20 people" after the Series A in May 2026. Launched 2024; still running.

AI agents that document, change, test and run the SAP systems of large companies · United States · AI-native · started small, grew after funding · still running · group: Developer tools and coding agents

- Outcome: Festo and KION named as SAP customers, May 2026 (evidence grade of the headline number: E; no revenue figure)
- Launch: 2024; months from launch to the outcome: 28
- Team: 3 co-founders (Emma Qian, Sam Yang, Prof. Dr. Alexander Zeier), founded 2024; Fortune reported "still just 20 people" after the Series A in May 2026
- Outside funding: Seed led by Accel and Series A led by Chemistry, with Conviction and SAP.iO; more than US$40M in total by May 2026 (Fortune and the company)
- Main factor: AI research team plus a co-inventor of SAP HANA
- Meters, 1 to 4: evidence strength 1, copyable pattern 1, capital needed 3
- Tags: AI-native, Small team, Funded, Running, B2B, Enterprise software

Links:

- Visit novaintelligence.com: https://www.novaintelligence.com/

**The product.** An AI platform for the teams that run SAP, the business software that large companies use for finance, supply chain and manufacturing. Its agents read a company's whole SAP landscape and keep its documentation current, check which custom code can be replaced by SAP's standard functions, write and test new code, help with the release and go-live, and trace the root cause of production errors. Sold to large enterprises through demos; no prices are published.

**Why it worked.** SAP has set 2030 as the deadline for customers to move to its newer S/4HANA platform, and decades of custom code make that move slow and expensive. The co-founders started from research on AI that writes code and moved to SAP because, in the company's words, every dollar spent on SAP software drives about seven dollars of services. A co-founder who co-invented SAP HANA gave the AI team the domain knowledge.

Evidence grades: A = disclosed by a public company, a filing or a payment processor, or verified by one; B = two or more independent reports agree, or a journalist confirmed it; C = a founder claim that no one else verified; D = reported, not verified: one founder claim or a third-party estimate with no second source; E = no revenue figure found, so the headline is a non-money fact and the other facts carry their own grades; calc = computed on this page from graded facts.

## Summary

Nova Intelligence sells AI agents, programs that use a large language model to plan a task, take steps in other systems and check the result, for SAP systems. SAP is the German vendor of ERP software (enterprise resource planning: one system for a company's finance, purchasing, production and sales). Large companies have changed their SAP systems over decades with custom code written in SAP's language ABAP, and must now move to S/4HANA, SAP's current platform, while keeping a Clean Core, SAP's term for a system with as few changes to the standard software as possible. Nova's agents read that code, document it, rewrite it and test it.

Fortune reported on 5 May 2026 that Emma Qian, a former research engineer at Google DeepMind and Meta AI, founded Nova with Sam Yang, earlier the founder of the investing app Bloom, and Prof. Dr. Alexander Zeier, a co-inventor of SAP HANA. The company's own story says Qian and Yang began with a research project on code generation and turned to SAP when they saw how much companies spend on SAP services. A directory gives 2024 as the founding year. Fortune reported a US$31.5 million Series A, more than US$40 million in total, and a team of 20 people.

No revenue figure for Nova was found, so the evidence grade is E. What is checked is the live product site, the founders, the funding as reported by Fortune and the company, and named customers: the CIO of Festo, a German automation company, is quoted on Nova's site, and KION Group is named in the funding announcement.

## Pricing

Nova Intelligence prices in November 2026: not published; the site offers only "Get a demo" (grade C)

| Plan | Per month, billed yearly (US$) | Per month, billed monthly (US$) | Limit |
|---|---|---|---|
| Enterprise platform |  |  | price not published; demo on request |

The site sells the platform to large enterprises through demos and contracts. No plan, seat price or usage price appears on any page read, so the table has one row.

## How much revenue does Nova Intelligence make? The numbers, graded

A three-founder AI company for SAP systems, founded in 2024, with more than US$40 million raised and about 20 people in May 2026; no revenue figure published

| Fact | Value | Grade | Note | Sources |
|---|---|---|---|---|
| Status, November 2026 | live, sold to enterprises through demos | C | Home page: "The Agentic Platform for Enterprise Systems", with "Get a demo", case studies and a list of trade events in November 2026 to January 2027. | [1] |
| Founders | Emma Qian (CEO), Sam Yang (CTO), Prof. Dr. Alexander Zeier (Chief Scientist) | C | Fortune: Qian "founded Nova alongside repeat operator Sam Yang and Professor Alexander Zeier". The company's posts call all three co-founders. | [3, 2, 4] |
| Founding year | 2024 | C | Simplify company profile: "Founded 2024". No company page read gives the year; the Internet Archive answered 429 Too Many Requests. | [9] |
| Team, May 2026 | 20 people | C | Fortune, 5 May 2026: "The company is still just 20 people." Simplify shows a size band of 1-10 and, in its own text, "the 20-person company". | [3, 9] |
| Funding | more than US$40 million, seed and Series A | C | Fortune reported a US$31.5 million Series A led by Chemistry, "bringing total funding to more than $40 million". The company says the seed was led by Accel, with Conviction and SAP.iO, SAP's venture arm, in both rounds. | [3, 2] |
| Named customers | Festo and KION Group | C | The funding announcement names both and quotes their CIOs. Festo's CIO is quoted on a page of the company's site; the quotations were not checked with Festo. | [2, 7] |
| Customer results claimed | more than 5x productivity at Festo; 2 years earlier private-cloud readiness | C | Festo case study: the CIO says the team estimates it is "more than 5x more productive on average" and moved its deadline from 2030 to the end of 2028. These are a customer's statements on the vendor's site. | [7] |
| Earlier case study | April 2025, with Kyndryl | C | Fortune: "An April 2025 case study with Kyndryl showed Nova's system led to a 75% reduction in manual effort and a 50% cost reduction". | [3] |
| AI model | Anthropic Claude through Amazon Bedrock | C | Sub-processors page: "Foundation model inference (Anthropic Claude) via Amazon Bedrock", in the customer's chosen processing region. | [8] |
| Programs and partners | Chevron Technology Ventures Catalyst Program, November 2025; strategic investment from SAP.iO | C | Company announcements dated 13 November 2025 (Chevron) and the SAP.iO post. | [10, 6] |
| Time from founding year to the May 2026 facts | about 28 months at most | calc | Calculated: January 2024, the earliest month of the founding year, to May 2026. | [9, 3] |

How to read this table: a grade C value is what the company states on its own pages, what a journalist reported from the company, or what a directory lists. A "calc" value (grade X) was computed on this page. No value comes from a payment processor, a filing or an audited report.

## When did Nova Intelligence launch, and how did it grow? Timeline

From a research project on AI code generation to named SAP customers and a Series A in about two years, read from the company's pages, Fortune and a jobs directory. Each entry carries its grade.

- **2024** (grade C [9, 4]): Founding year in the Simplify directory. Qian and Yang begin with a research project on code generation, then turn to SAP; Zeier joins as Chief Scientist and co-founder.
- **April 2025** (grade C [3]): A case study with Kyndryl reports 75% less manual effort and 50% lower cost for an SAP modernization, Fortune says.
- **13 November 2025** (grade C [10]): Selected for the Chevron Technology Ventures Catalyst Program.
- **5 May 2026** (grade C [3, 2]): Fortune reports a US$31.5 million Series A led by Chemistry, more than US$40 million in total, and a team of 20. The company names Festo and KION as customers.
- **November 2026** (grade C [1, 7, 8]): The site shows customer case studies, a sub-processor list with Anthropic Claude on Amazon Bedrock, and open jobs.

## Why did Nova Intelligence work?

### AI research team plus a co-inventor of SAP HANA

Two co-founders came from AI research and consumer apps; the third led SAP's HANA project from 2006 to 2012 and was CTO of Accenture's SAP Business Group, he writes. The combination gave a young company access to the CIOs of large SAP customers.

### A fixed deadline creates demand

SAP's 2030 deadline for the move to S/4HANA forces customers to understand and reduce decades of custom code. Nova sells into that project and then into daily operations, where its agents triage tickets and fix errors.

### Knowledge kept inside the platform

The company says each customer's knowledge of its own system builds up in the platform, which it calls the Enterprise Knowledge Core, instead of staying with a few senior staff or outside consultants.

## How it was built

These details come from the company's pages. They are grade C.

- Foundation models: Anthropic Claude through Amazon Bedrock, run in the processing region each customer chooses.
- Hosting on Amazon Web Services: ECS, RDS, EC2, S3 and CloudWatch, as the sub-processors page lists.
- Tailscale for private networking between the company's servers, Firecrawl for web search, and Pylon for customer support.
- The agents write code changes into the SAP system, run them and test them, with approvals where the customer requires them, the home page says.

Sources: [8, 1]

## Who tried this before Nova Intelligence, and why did they fail?

Four companies in a database of failed startups promised to make large companies' software easier to understand, connect or change. The values below are from that database and are not independently verified.

The failed startups come from our database of dead startups, retold here in our own words; Its entries are AI-assisted summaries of public sources, so they carry no evidence grade.

### CodeSee (2020 to 2024; funding US$10M; cause: No Market Need)

- Promise: Help developers understand complex codebases faster with interactive maps of their architecture.
- Why it died: It could not prove a return to engineering leaders, and free tools covered the same need.
- Record: 

### Solid (2017 to 2025; funding US$81M; cause: Unit Economics)

- Promise: One API to connect enterprise systems without custom integration work.
- Why it died: Enterprise integration was harder than expected; long sales cycles and support costs broke the unit economics.
- Record: 

### Primary Data (2013 to 2018; funding US$100M; cause: Product/Tech Failure)

- Promise: Let enterprises reach data across all storage systems without migrations.
- Why it died: It spent its money on edge cases while the market moved to simple cloud migration tools.
- Record: 

### Sider (2012 to 2024; funding US$7M; cause: Competition)

- Promise: Automated code review that catches bugs and security flaws for development teams.
- Why it died: Free open-source tools and large platforms that bundled code analysis took its market.
- Record: 

## What changed

CodeSee and Sider sold insight into code, and buyers saw it as nice to have. Solid and Primary Data sold one layer over many enterprise systems and drowned in the edge cases of each one. Nova's agents do the work itself, writing, testing and fixing code, which is easier to price against consultant hours than a map or a report.

Two things differ. Large language models can now read old custom code and write new code, and SAP's 2030 deadline gives the buyer a fixed date and a budget. Nova also narrowed to one vendor's system, SAP, instead of promising to cover every enterprise system at once.

## Can one person copy Nova Intelligence now?

**Verdict: Not by one person.**

Selling to the CIOs of large SAP customers took a co-founder with thirty years at SAP and Accenture, security reviews, and more than US$40 million in funding, as Fortune reported. The pattern that a small founder can use is narrower: pick one large legacy system with a deadline and build AI that does the migration work, not only advice about it.

1. Find a legacy platform with a fixed migration deadline and a large services market around it.
2. Bring in a co-founder or adviser whom that platform's customers already trust.
3. Let the agent finish work in the customer's system, with approvals, instead of producing drafts.
4. Publish named customer case studies with a quoted CIO.

**What is still open.** Smaller legacy systems with the same pattern: older ERP products for mid-sized companies, industry systems in insurance or hospitals, and migrations forced by a vendor's end-of-support date.

## What could go wrong for Nova Intelligence?

- **What is not verified.** No revenue figure for Nova was found. What was looked for: a payment-processor profile, a founder interview with numbers, a filing, and press figures; Fortune and the company give funding and team size but no revenue. A directory card that suggested US$3 million ARR for Nova attaches that figure, read on 6 November 2026, to Bloom, a different product, so Arrfounder's figure is not used. The founding year 2024 rests on one directory. A reader could look for a revenue figure in later press about the company.
- **Customer results are the vendor's pages.** The 5x productivity, the KION results and the Festo deadline come from case studies on Nova's own site and from Fortune quoting the CEO. They are customer and vendor statements, not audited outcomes, and are not revenue.
- **Dependence on one model provider and on SAP.** The sub-processors page lists Anthropic Claude through Amazon Bedrock as the foundation model. Simplify's summary notes that SAP's own Joule for Developers offers ABAP migration and testing automation; Fortune asked the CEO about SAP building more of this itself, and she answered that Nova is complementary.
- **Competition on this site.** No other story on this site sells AI for SAP systems; the closest are other developer-tools stories about AI coding agents, which serve general software teams.
- **A small start that grew.** Three co-founders started the company; Fortune reported 20 people and more than US$40 million raised in May 2026, and the company is hiring in engineering, sales and marketing.

## Sources

1. Nova Intelligence home page, read 6 November 2026 — https://www.novaintelligence.com/ (grade C). Read in full. Source for the product, the demo-only sales and the events list.
2. Nova Intelligence blog: Nova Intelligence Raises $40M to Build the Frontier AI Platform for SAP — https://www.novaintelligence.com/blog/series-a (grade C). Read in full. Source for the rounds, the investors, the co-founders and the named customers Festo and KION.
3. Fortune, 5 May 2026: Exclusive: Nova Intelligence raises $31.5 million to bring agentic AI to SAP's $89 billion migration wave — https://fortune.com/2026/05/05/exclusive-nova-intelligence-ai-sap-chemistry-emma-qian/ (grade C). Read in full. Source for the founders, the Series A, the total funding, the team of 20 and the Kyndryl case study.
4. Nova Intelligence blog: From General Coding Agents to SAP: Founding Nova Intelligence — https://www.novaintelligence.com/blog/from-general-coding-agents-to-sap (grade C). Read in full. Source for the founding story.
5. Nova Intelligence blog, by Prof. Dr. Alexander Zeier: SAP to AI: HANA Co-Inventor's Next Bet on AI — https://www.novaintelligence.com/blog/co-inventing-hana-to-co-founding-nova-ai-my-next-big-bet (grade C). Read in full. Source for the co-founder's SAP background.
6. Nova Intelligence blog: Nova Intelligence Announces Strategic Investment from SAP iO Fund — https://www.novaintelligence.com/blog/saps-strategic-investment-in-nova-intelligence (grade C). Read in full. Source for the SAP.iO investment and the founding team.
7. Nova Intelligence case study on Festo: Clean Core program brought forward by 2 years — https://www.novaintelligence.com/customers/festo (grade C). Read the first sections. Source for the Festo quotations and claimed results.
8. Nova Intelligence sub-processors page — https://www.novaintelligence.com/subprocessors (grade C). Read in full. Source for the model provider and the infrastructure.
9. Simplify company profile: Nova Intelligence — https://simplify.jobs/c/Nova-Intelligence (grade C). Read the company facts and summary. A jobs directory. Source for the founding year and the critics' note on SAP Joule.
10. Nova Intelligence blog: Nova Intelligence Selected for Chevron Technology Ventures Catalyst Program, 13 November 2025 — https://www.novaintelligence.com/blog/nova-intelligence-selected-for-chevron-technology-ventures-catalyst-program (grade C). Read the opening. Source for the date.
11. Database of failed startups, local export — /method/#dead (grade C). Local export. AI-written summaries, not independently verified. Source for the failed predecessors.

Contact: aiolum@livyden.com. Run Nova Intelligence? Send corrections or a verified figure to aiolum@livyden.com (subject: Correction: Nova Intelligence).