DeepTempo
Adds an intelligent prediction and detection layer to security stacks using AI-powered LogLM for threat detection.
Website: https://www.deeptempo.ai
Cover Block
| Field | Value |
|---|---|
| Name | DeepTempo |
| Tagline | Adds an intelligent prediction and detection layer to security stacks using AI-powered LogLM for threat detection |
| Headquarters | Palo Alto, California |
| Founded | 2023 |
| Stage | Pre-Seed |
| Business Model | SaaS |
| Industry | Cybersecurity |
| Technology Type | AI / Machine Learning (foundation model for security logs) |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Solo Founder (Evan Powell) |
| Accelerator | BNY Ascent Program |
Links
- Website: https://www.deeptempo.ai
- LinkedIn: https://www.linkedin.com/company/deeptempo
- Medium (engineering and product blog): https://medium.com/deeptempo
- PitchBook profile: https://pitchbook.com/profiles/company/753291-19
The Short Version
DeepTempo is a Palo Alto-based cybersecurity startup building what it calls a LogLM, a vertical foundation model trained on security log data, that runs natively inside Snowflake to detect attack patterns that traditional network detection and SIEM tools miss [DeepTempo website, retrieved 2025]. The company was founded in November 2023 by Evan Powell, a repeat infrastructure-software founder whose prior companies include Nexenta Systems, StackStorm (sold in 2017), and Reprise, and emerged from stealth in November 2024 with the launch of its Tempo app on the Snowflake Marketplace [Blocks and Files, Feb 2025] [Blocks and Files, Nov 2024]. The product positions itself as a detection and triage layer that maps anomalies to MITRE ATT&CK patterns, surfaces likely impacted entities, and reduces analyst load on overloaded SOC teams [The Fast Mode, retrieved 2026]. Early validation includes participation in the BNY Ascent Program in collaboration with BNY Mellon, with BNY also cited by the founder as a design partner on the LogLM itself [Fintech Futures, retrieved 2026] [LinkedIn]. No priced funding round has been disclosed publicly. In Q1 2025 the company hired Chris Bowen, formerly SVP of sales at Hammerspace, as its first VP of Sales [Blocks and Files, Feb 2025]. Over the next 12 to 18 months, the questions worth tracking are whether DeepTempo can convert Snowflake Marketplace distribution into paying enterprise customers, whether the LogLM approach holds up against well-capitalized incumbents in cloud-native detection, and whether a priced seed round materializes to fund the sales build-out.
Data Accuracy: GREEN -- Confirmed by Blocks and Files, DeepTempo website, LinkedIn, and Fintech Futures.
Taxonomy Snapshot
| Axis | Value |
|---|---|
| Stage | Pre-Seed |
| Business Model | SaaS (Snowflake Native App) |
| Industry / Vertical | Cybersecurity, threat detection |
| Technology Type | Deep learning foundation model (LogLM) |
| Geography | North America (Palo Alto, CA) |
| Growth Profile | Venture Scale |
| Founding Team | Solo Founder, repeat operator |
The Company in Brief
DeepTempo was set up by Evan Powell in November 2023 and operated in stealth for roughly twelve months before publicly launching its Tempo app in November 2024 [Blocks and Files, Feb 2025] [Blocks and Files, Nov 2024]. The thesis was "to fix cyber, from the foundations," with the first concrete output being a LogLM (log-data language model) developed with BNY and other unnamed design partners [LinkedIn]. The company is headquartered in Palo Alto, California.
The public milestone sequence is short but coherent. Founding in late 2023 was followed by a stealth-period buildout of the LogLM and the Snowflake Native App. The November 2024 launch on the Snowflake Marketplace was paired with a Snowflake Startup Spotlight write-up [Snowflake]. In February 2025 the company hired its first VP of Sales, Chris Bowen, previously SVP of sales at Hammerspace [Blocks and Files, Feb 2025]. Participation in the BNY Ascent Program is the only named program affiliation surfaced in the public record [Fintech Futures, retrieved 2026].
No priced funding round, named lead investor, customer count, or revenue figure has been disclosed. PitchBook lists a profile but the underlying valuation and cap-table fields are not visible without a subscription [PitchBook].
Data Accuracy: GREEN -- Confirmed by Blocks and Files, Snowflake, LinkedIn, and Fintech Futures.
What They Have Built
Tempo, the company's flagship product, is delivered as a Snowflake Native App that ingests security log data already resident in a customer's Snowflake account and applies deep learning to flag likely security incidents [Blocks and Files, Nov 2024]. The pitch is additive: DeepTempo positions LogLM as "an intelligent prediction and detection layer" on top of existing NDR and SIEM stacks [DeepTempo website, retrieved 2025].
The technical center of gravity is the LogLM itself, described by the company as a vertical foundation model purpose-built for security logs [DeepTempo website, retrieved 2025]. Tempo enriches anomalies with context including similar attack patterns drawn from the MITRE ATT&CK matrix, lists of potentially impacted entities, and triage-ready metadata [The Fast Mode, retrieved 2026]. The Snowflake-native architecture means data does not have to be exfiltrated to a vendor cloud.
No public roadmap, model card, benchmark result, or third-party evaluation has been published. The strongest external validation point on the technology side is BNY Mellon's role as a named design partner on the LogLM [LinkedIn].
Data Accuracy: YELLOW -- Product claims confirmed by company website, Blocks and Files, and Snowflake; technical depth and benchmarks not independently verified.
Market Research and Opportunity
Security operations is a category where buyer pain has outpaced incumbent tooling, driving a wave of AI-native detection startups since 2023. DeepTempo is targeting the detection-and-response layer of the security operations center, a category historically dominated by SIEM platforms (Splunk, IBM QRadar, Microsoft Sentinel) and newer cloud-native SIEMs [Blocks and Files, Nov 2024].
Enterprises are consolidating logs into cloud data warehouses (Snowflake, Databricks, BigQuery) for cost and analytics reasons, which creates a structural opening for detection products that run where the data already lives. Tempo's distribution choice (Snowflake Native App, listed on Snowflake Marketplace) is directly aligned with that shift [Snowflake] [Blocks and Files, Nov 2024].
The relevant tailwinds are three. First, the Snowflake Native App framework itself. Second, regulated-industry interest in keeping security telemetry inside the customer's own cloud account for data-residency reasons. Third, the willingness of a Tier-1 financial institution (BNY Mellon) to act as a design partner and accelerator host [Fintech Futures, retrieved 2026] [LinkedIn].
Adjacent and substitute markets to monitor include managed detection and response (MDR) services, cloud detection and response (CDR) tools from incumbents like Wiz and CrowdStrike, and in-warehouse security analytics offerings that Snowflake or Databricks could build natively.
| Market signal | Detail | Source |
|---|---|---|
| Distribution channel | Snowflake Marketplace listing as Native App | [Snowflake] [Blocks and Files, Nov 2024] |
| Named design partner | BNY Mellon (also accelerator host) | [LinkedIn] [Fintech Futures, retrieved 2026] |
Data Accuracy: YELLOW -- Channel and partner signals confirmed by Snowflake, Blocks and Files, Fintech Futures, and LinkedIn; no third-party TAM cited.
Who Else Is Fighting for This
DeepTempo sits inside a heavily contested category, and its differentiation rests on a specific architectural bet (foundation model trained on logs, deployed inside the customer's Snowflake account).
| Company | Positioning | Stage / Funding | Notable Differentiator |
|---|---|---|---|
| DeepTempo | LogLM-based detection layer running as a Snowflake Native App | Pre-seed, no priced round disclosed | Vertical foundation model on security logs, in-warehouse deployment |
| Wiz | Cloud security platform (CNAPP) with detection capabilities | Late-stage, well-capitalized public-market-adjacent | Breadth of cloud posture, vulnerability and detection coverage |
Incumbent SIEM vendors (Splunk, Microsoft Sentinel, IBM QRadar) own the buying center but are viewed as expensive at scale. Cloud-native detection challengers (Wiz, CrowdStrike, SentinelOne) have strong distribution and growing detection breadth. Adjacent substitutes include in-warehouse analytics offerings that Snowflake or Databricks customers can build themselves.
Where DeepTempo has a defensible edge: the Snowflake Native App distribution model lowers the data-movement and procurement friction that slows down most security software sales [Snowflake]. The founder's prior infrastructure-software exits (Nexenta, StackStorm, Reprise) and the BNY design-partner relationship give the company credibility [TechCrunch, Feb 2013] [DeepTempo blog, Jun 2025] [LinkedIn].
Where the company is most exposed: Wiz and the larger cloud-security incumbents have direct relationships with the CISO buying center and the budget gravity that comes with multi-product platforms. The renewal motion at enterprise ACV is unproven, the sales team is one VP-level hire deep [Blocks and Files, Feb 2025], and the public record does not yet show a published model benchmark.
Opportunity
If DeepTempo's LogLM thesis holds, the upside is to become the default AI detection layer for any enterprise that already runs security telemetry through a cloud data warehouse.
The headline opportunity. The most ambitious credible outcome for DeepTempo is to become the embedded detection model for warehouse-resident security data. Tempo is already shipping as a Snowflake Native App with Snowflake's own promotional support [Snowflake], a Tier-1 financial institution acted as a design partner on the underlying model [LinkedIn], and the founder has previously taken open-source infrastructure projects (StackStorm, OpenEBS) into commercial outcomes [DeepTempo blog, Jun 2025] [Kubernetes Podcast, retrieved 2026].
Growth scenarios.
| Scenario | What happens | Catalyst |
|---|---|---|
| Snowflake-channel land-and-expand | Tempo becomes a top-installed security app on the Snowflake Marketplace | Snowflake co-marketing plus a published detection benchmark |
| Financial-services beachhead | DeepTempo wins two to three named bank logos on the back of the BNY relationship | Public case study from BNY Mellon or a peer institution |
| Multi-warehouse expansion | The LogLM is ported to Databricks and BigQuery | A second native-app listing (Databricks Marketplace) |
What compounding looks like. The flywheel is a data flywheel: each additional design-partner deployment generates more labeled attack and benign-log examples that improve the LogLM, which raises detection quality. The distribution flywheel is channel-driven: every new Snowflake Marketplace install is a low-friction trial.
Data Accuracy: YELLOW -- Scenario inputs confirmed by Snowflake, LinkedIn, Fintech Futures, and DeepTempo blog; valuation comparisons are scenario analysis.
Sources
- [Snowflake] Snowflake Startup Spotlight: DeepTempo | https://www.snowflake.com/en/blog/startup-spotlight-deeptempo-cybersecurity-ai/
- [DeepTempo website, retrieved 2025] DeepTempo, Close the Detection Gap | https://www.deeptempo.ai
- [LinkedIn] DeepTempo company page | https://www.linkedin.com/company/deeptempo
- [LinkedIn] Evan Powell, Many time founder and 5 exits | https://www.linkedin.com/in/epowell/
- [Blocks and Files, Feb 2025] DeepTempo hires first sales VP to pitch AI log security | https://blocksandfiles.com/2025/02/18/deeptempo-hires-hammerspace-sales-boss-to-sell-its-ai-driven-snowflake-security-sw/
- [Blocks and Files, Nov 2024] DeepTempo unveils AI-powered app for detecting security incidents | https://blocksandfiles.com/2024/11/12/deeptempo-security-log-data/
- [PitchBook] DeepTempo 2026 Company Profile | https://pitchbook.com/profiles/company/753291-19
- [Medium] DeepTempo publication | https://medium.com/deeptempo
- [TechCrunch, Feb 2013] Data Storage Software Company Nexenta Systems Raises $24M | https://techcrunch.com/2013/02/27/data-storage-software-company-nexenta-systems-raises-24m/
- [Bloomberg Markets] Evan Powell, Cloudbyte Inc, Profile and Biography | https://www.bloomberg.com/profile/person/21520018
- [Bloomberg Markets] Evan Powell, Reprise Inc, Profile and Biography | https://www.bloomberg.com/profile/person/25083177
- [Forbes Technology Council] Evan Powell, Co-Founder, Reprise | https://councils.forbes.com/profile/Evan-Powell-Co-Founder-Reprise/1974c2b9-01d8-4d03-8544-c9ca3394d051
- [Forbes] Evan Powell, Forbes Technology Council profile | https://www.forbes.com/councils/forbestechcouncil/people/evanpowell/
- [Forbes, Feb 2012] Nexenta Aims At EMC's Heart | https://www.forbes.com/sites/petercohan/2012/02/16/nexenta-aims-at-emcs-heart/
- [Forbes, Jan 2024] Three Enterprise Tech Presales Trends To Navigate For 2024 | https://www.forbes.com/councils/forbestechcouncil/2024/01/30/three-enterprise-tech-presales-trends-to-navigate-for-2024/
- [Fintech Futures, retrieved 2026] BNY Ascent Program coverage referencing DeepTempo participation | https://www.fintechfutures.com
- [The Fast Mode, retrieved 2026] DeepTempo Tempo app coverage | https://www.thefastmode.com
- [DeepTempo blog, Jun 2025] Founder reflections including StackStorm history | https://medium.com/deeptempo
- [Kubernetes Podcast, retrieved 2026] Episode featuring Evan Powell on OpenEBS and cloud-native storage | https://kubernetespodcast.com
Articles about DeepTempo
- DeepTempo Wants Snowflake to Be the Place Security Teams Catch the Attack SIEM Missed — Evan Powell's pre-seed bet: a vertical foundation model for log data, sold as a Native App inside the warehouse customers already run.