Edelbyte launches with $70M for AI-agent data analytics | Magica
Edelbyte launches with $70M to tackle AI agents’ data costs
Editorial Team
••📖6 min read
The startup founded by Microsoft data-system veterans has raised $70 million across seed and Series A rounds. Its unified analytics platform targets unpredictable AI-agent workloads, but the claimed savings still need comparable performance evidence.
Edelbyte emerged from stealth with $70 million across two rounds; its latest financing reportedly values it at $500 million.
Its four founders bring long Microsoft data-systems experience, but its customer and efficiency disclosures remain company claims.
Existing rivals already combine some search and analytics functions, making total operating cost a key test.
Edelbyte, a search and analytics startup founded by four former Microsoft data engineers, emerged from stealth on November 11 with $70 million in seed and Series A funding. The company is pitching a single data engine to enterprises facing more frequent and less predictable queries from AI agents.
The financing comprises a $25 million seed round led by venture investor Sequoia Capital and a $45 million Series A led by Quiet Capital. The latest round values the company at $500 million, according to the financing report—roughly a year after its November 2025 founding.
That backing finances a challenge to data-platform suppliers including Snowflake, Databricks and ClickHouse. Edelbyte's argument is that agents generating many complex questions will make the cost of each query more consequential. Its launch announcement says some customers migrating from legacy systems can already run ten times as many queries within the same budget. The evidence disclosed so far does not establish a comparable advantage across competitors or workloads.
Edelbyte raised a $25 million seed round followed by a $45 million Series A. Chart: Magica; data: CTech. Source: CTech.
Microsoft experience behind a new engine
Co-founder and CEO Avner Aharoni previously helped develop Microsoft's Kusto data platform. His fellow founders, chief technology officer Evgeney Ryzhyk and engineering vice president Alexander Sloutsky, were members of Kusto's original team; Aharoni and platform-engineering vice president Yoni Leibowitz joined later. Kusto began as an internal Microsoft project and became the foundation for products including Azure Data Explorer, a managed service for analyzing large volumes of real-time and time-series data.
Each founder has approximately 20 years of Microsoft experience, and the four spent more than a decade working together on large data systems, the . That is relevant experience for building a database engine, although it cannot establish the performance of the new product.
Treasury Secretary Scott Bessent said agencies had backed export restrictions on Nvidia customer Megaspeed for months. He challenged Commerce’s handling of the case and Nvidia’s safeguards, while Megaspeed denies illicit chip exports to China.
In a launch interview, Aharoni said the team reconsidered implementation after its work on Kusto, seeking a faster, cheaper platform with a unified engine and query federation—the ability to query across data sources. He singled out CPU efficiency as important when agents generate more requests.
Edelbyte has approximately 40 employees across Israel, Europe and the United States, with offices in Tel Aviv and San Francisco. Greenoaks and Swish Ventures joined the seed round; all existing investors participated in the Series A, alongside new investors including QP Ventures, Startpoint Capital and Cerca Partners, according to the financing report.
The company says it operates petabyte-scale systems for Fortune 500 companies and other organizations in cybersecurity, automotive manufacturing and financial services. Aharoni separately described four significant design partners in production environments at petabyte scale in the launch interview. These accounts establish what Edelbyte says about early use, without disclosing paying-customer numbers or revenue.
What the cost promise covers
Edelbyte says its platform combines traditional analytics, text search, spatial queries and AI workloads. Its proposed economy comes from replacing separate systems and reducing replicated data, as well as compressing the information stored and processed for queries.
Those are company claims, as are the millisecond response times and support for thousands of concurrent queries described in the product coverage. Edelbyte says it indexes all available data so users can ask ad hoc questions, including questions about extensive historical records, rather than relying on indexes prepared for particular dashboards, reports or alerts. The coverage supplies no benchmark connecting those latency and concurrency claims to a specified workload.
The platform also offers managed and self-hosted deployment across public clouds, private clouds and on-premises infrastructure, according to the company's announcement. That broadens the pitch to enterprises with different requirements for where their data systems run.
The announcement's ten-times-as-many-queries claim applies to some customers moving from unspecified legacy systems. It names neither those customers nor the systems being replaced and supplies no reproducible query mix, hardware configuration or cost breakdown. It is a claim about query volume within a budget, rather than evidence that every query runs ten times faster or that every customer can cut its bill by the same proportion.
Rivals narrow the differentiation
Databricks, an established enterprise data-platform supplier, already offers Agent Bricks for building agents grounded in enterprise data, as the launch coverage notes. Agent support alone therefore does not distinguish Edelbyte.
There is also a challenge beyond query speed. In that coverage, Andy Neill, vice president of the technical counsellor program at technology research and advisory firm Infotech, identified agents' ability to consume data and context consistently as a major obstacle. Faster responses would not by themselves establish that an agent can use the returned information reliably.
ClickHouse, the analytical database provider Edelbyte names as a competitor, announced native full-text search's general availability on March 10, 2026. It uses inverted indexes, mapping text tokens to rows, to combine text filtering with analytical aggregation. That is an existing alternative for workloads needing both functions.
The scope matters: ClickHouse says the feature does not provide relevance scoring or directly accelerate phrase searches. It is designed for token filtering and analytics, with dedicated search engines remaining an alternative for richer ranking and linguistic requirements. This overlap does not establish that ClickHouse and Edelbyte cover identical workloads.
ClickHouse's announcement also makes an economic tradeoff explicit. In its own benchmark on 50 terabytes of log data, using 59 virtual CPUs and 256 gigabytes of RAM in ClickHouse Cloud, adding the text index reduced insert throughput by approximately 50% compared with no index. The company also says text indexes consume more storage than its earlier Bloom filter indexes. Those findings concern ClickHouse's implementation, not Edelbyte's, but show why an efficiency comparison needs to include ingestion and storage alongside query response times.
Company-reported ClickHouse index storage on its GitHub Events dataset, separate from the 50 TB log benchmark. Chart: Magica; data: ClickHouse. Source: ClickHouse.
What buyers need to see next
Edelbyte says its next priorities are expanding its enterprise customer base, international presence and team. Its funding and founders' experience give it resources and relevant credentials for that effort; the purchasing question still requires workload-specific evidence.
A useful comparison would hold data, query mix, concurrency and service requirements constant, then disclose response times and total spending. It would also account for data ingestion, index construction and maintenance, storage, and the work required to migrate existing systems. The launch disclosures do not provide that comparison.
For prospective customers, the unresolved question is whether consolidating systems into Edelbyte reduces the full cost of their actual workload while preserving the performance and data access their applications need. Named deployments and reproducible cost results would make that decision more concrete than the launch's broad efficiency claims.
Applied Materials expects chipmaking equipment demand to rise more than 30% in 2027, while supplier capacity and factory readiness will determine how much demand becomes sales.
Shi Bingzhe and two companies face money laundering charges over 917 servers allegedly acquired from Aperia International. The case now covers five individuals and six companies, but the new charges do not establish the servers’ chip contents or ultimate destination.
Nothing’s December AI device has appeared with red and green lights and an apparent recording button. The clues suggest an audio tool, but existing AI recorders leave the company needing to explain its hardware, service costs and use of personal context.
Apify says an intruder accessed customer accounts and personal information or stored data. Customer notices describe credential rotations and two-factor resets, while the affected-account count and full data exposure remain unclear.
Chinese computing projects still specify Nvidia processors, but their reported values mix proposals and contracts. Conditional US licenses, Chinese restrictions, financing and customer demand complicate the path to usable capacity.
The former Huawei server business has passed Shenzhen's listing committee review for an 8 billion yuan IPO. Its growing market share and profits sit alongside negative operating cash flow and a sharp increase in short-term borrowing.
EU leaders gather for their first dedicated AI discussion, weighing investment, foreign dependencies and safety while Ukraine diplomacy competes for attention. The informal summit is a strategic discussion, with no political statement scheduled for adoption.
New hardware orders have increased Nscale's 2027 funding needs and limited its ability to defer spending. The British AI infrastructure provider is pursuing a New York listing targeting a $35 billion valuation while financing for its Anthropic contracts remains uncommitted.
Commerce Secretary Howard Lutnick has agreed to voluntary Senate testimony as bipartisan scrutiny of Nvidia exports turns to government approvals and enforcement. A hearing date remains unconfirmed, and intelligence suspicions remain disputed.
Samsung Electronics is reportedly in talks to buy Modine as it expands a cooling business already bolstered by FläktGroup. Modine brings fast-growing data center sales, but weaker margins and a Texas factory still ramping production complicate the potential deal.
Nvidia faced pressure to enlarge its financing backstop before Jensen Huang departed. The leadership change raises questions about support for AI customers, but interim CEO Colette Kress helped build that strategy and chip demand remains a counterweight.
Google launched Gemini 4 Hydrogen and automatic model switching in Antigravity, claiming near-Carbon task success with less quota and faster completion. Lower token rates support the efficiency pitch, but reasoning settings, mixed competitive results and inconsistent benchmark labels leave the cost of completed work to be tested.
Japan is considering safety and reporting duties backed by enforceable orders for advanced AI. The proposal would strengthen its nonbinding framework, but capability thresholds, covered businesses and penalties remain unresolved.
Google says its new fast AI model is reaching Search users worldwide, improving complex answers and source discovery. The rollout expands access to Gemini 4, but its effect on answer quality and website traffic remains unmeasured.
Four nonprofits are asking a California court to stop five AI developers from building or deploying more capable systems. The case seeks enforceable limits even as OpenAI has previously reported a training pause and is launching defensive AI tools.
Expo Simulators gives coding agents remote iOS and Android environments with recordings for reviewers. The launch extends Expo's cloud development services into verification, with per-minute charges and shared capacity limits.
Kleiner Perkins and Clearlake Capital led Mirendil’s $1 billion Series A. Its agent-led model-building experiment points to smaller research teams, but company benchmarks, substantial compute costs and a private beta qualify the promise of wider access.
Mount Thor raised $45 million in equity, alongside separate hardware financing, to expand Apple infrastructure for AI agents and model inference. Existing Mac clouds and workload-specific performance limits put its demand and cost claims to the test.
Salesforce Ventures led a Series C that values Q-CTRL above $1 billion. The funding backs GPS-free navigation and quantum computing software, while undisclosed revenue and contested performance benchmarks leave the commercial case open.
