Pioneering the future of work with hyperautomation and Agentic AI for Cybersecurity & IT teams. 🚀 Insights on cybersecurity, AI, and automation. 🔔
Paris, France
Joined October 2021
- Tweets489
- Following156
- Followers265
- Likes632
If your remediation workflows still depend on analysts stitching actions together, the issue is not detection. It’s integration depth.
Endpoint telemetry, threat intelligence, cloud inventory, and operational insights all live in different systems.
When integrations are shallow, automation stops at investigation.
You know what needs to be fixed. You can't execute the response end to end.
Remediation only scales when containment, enrichment, validation, and recovery actions are fully executable through APIs.
Full read and write coverage. Not alerts. Not manual playbooks.
CrowdStrike Falcon · Imperva ThreatRadar · Sophos Intercept X · GCP Analytics Reporting · AWS Application Discovery Service
Tomorrow, 11:30 CEST.
45 minutes on the unglamorous half of vulnerability management: everything that happens after the scanner tells you what's wrong.
Can't believe you haven't registered yet
🔗 app.livestorm.co/hackuity/cl…
If your threat intelligence workflows still stop at collecting indicators, the issue is not intelligence. It’s integration depth.
Threat context, account data, threat streams, file integrity, and IOCs often live in separate systems.
When integrations are shallow, automation stops at enrichment.
You can identify the threat. You still need humans to connect the dots and take action.
Threat intelligence only scales when enrichment, investigation, validation, and response actions are fully accessible through APIs.
Full read and write coverage. Not dashboards. Not exports. Not manual handoffs.
Exabeam Context · Databricks Accounts · Anomali ThreatStream · Acronis Notary · IOC Parser
If your threat intelligence workflows still stop at collecting indicators, the issue is not intelligence. It’s integration depth.
Alerts, threat data, security events, cloud data, phishing signals. They live across different systems.
When integrations are shallow, automation stops at enrichment.
You have the intelligence. You still need humans to connect the dots and act.
Threat intelligence only scales when detection, enrichment, correlation, and response actions are fully executable through APIs.
Full read and write coverage. Not dashboards. Not feeds. Not manual handoffs.
PagerDuty Events · Trellix MVISION · Kaspersky · Imperva Cloud Data Security · Fortra Agari
📅 Tuesday, September 15, 2026 · 11:30 am CEST
Sylvain Cortes (VP Strategy, Hackuity) and Umberto MATTEI (Mindflow) go hands-on for 45 minutes: prioritizing what actually matters in your environment, then automating the response at scale with agentic AI.
Real workflows, built by people running them in production. Adapt them for your own stack.
🔗 app.livestorm.co/hackuity/cl…
If your endpoint protection workflows stop at alerts, the issue is not detection. It’s integration depth.
EDR signals, antivirus events, file integrity, network controls, phishing data. They live in different systems.
When integrations are shallow, automation stops at isolation.
You can flag the endpoint. You can’t coordinate the full response.
Endpoint protection only scales when containment, remediation, validation, and follow-up actions are fully executable through APIs.
Full read and write coverage. Not alerts. Not manual playbooks.
Trellix MVISION EDR · Kaspersky · Acronis Notary · pfSense · Fortra Agari
20 minutes of tab-juggling, or a few seconds.
In our webinar with Siit, Umberto MATTEI walked through what changes when ticket triage runs on agentic AI. The time saving is the obvious part: nobody's copy-pasting analysis results between tools and then updating the ticket by hand.
The less obvious part is consistency. Every team handles incidents differently. Procedures drift, conclusions vary, and the same ticket gets two different answers depending on who picked it up.
An automated workflow runs the same way every time and leaves an audit trail. It also means fewer tickets are escalated before anyone checks whether they need to be.
Full session → youtu.be/4FE8d0GPZ8o?si=SdVo…
If your cloud security workflows still rely on switching between consoles, the issue is not visibility. It’s integration depth.
Traffic controls, access policies, threat signals, logs, and behavior analytics. They live across different platforms.
When integrations are shallow, automation stops at detection.
You can see the anomaly. You can’t enforce the response.
Cloud security only scales when policy updates, traffic controls, detections, and response actions are fully executable through APIs.
Full read and write coverage. Not dashboards. Not alerts alone.
Netskope · Zscaler · Imperva SIEM · Logpoint XDR · Lacework Polygraph
If your application management still depends on manual approvals and fragmented tooling, the issue is not process. It’s integration depth.
App connectivity, service tickets, API gateways, service mesh configs, data streams. They live in different systems.
When integrations are shallow, automation stops at provisioning.
You can request access or changes. You can’t execute them end-to-end.
Application management only scales when policy enforcement, service configuration, API control, and data flow operations are fully executable through APIs.
Full read and write coverage. Not tickets. Not manual handoffs.
Tufin SecureApp · Ivanti Cherwell · Azure API Management · AWS App Mesh · Lenses
If your credential management still relies on manual reviews and periodic cleanups, the issue is not policy. It’s integration depth.
Tokens, identities, user permissions, service accounts, access relationships. They live across multiple systems.
When integrations are shallow, automation stops at visibility; you can audit access. However, you can’t enforce it in real time.
Credential management only scales when provisioning, deprovisioning, token rotation, and access updates are fully executable through APIs.
Full read and write coverage, not audits or spreadsheets.
@databricks Token Management · @Imperva User Management · @Microsoft Graph Identity · @Acronis Account Management · Veza
Nice to spot Mindflow in the 2026 Digital Workplace Radar by Wavestone, in the Process & Workflow Automation category.
The framing they use is one we like: no-code orchestration that connects the many tools a company already runs, automates the workflows between them, and quietly reduces the operational load on IT and security teams.
Less about adding another platform, more about removing the manual work sitting between the ones you have.
The radar maps 108 French startups reshaping the digital workplace, a good read if you're curious where day-to-day work is heading. Thanks to the team behind it.
If your remediation workflows still depend on analysts stitching actions together, the issue is not detection. It’s integration depth.
Endpoint telemetry, threat intelligence, cloud inventory, and operational insights all live in different systems.
When integrations are shallow, automation stops at the investigation stage.
You know what needs to be fixed. You can't execute the response end to end.
Remediation only scales when containment, enrichment, validation, and recovery actions are fully executable through APIs.
Full read and write coverage, not alerts, not manual playbooks.
"Securing AI" is now its own category on the map, and that says a lot about where things are heading.
Mindflow was featured in the 2026 French Cybersecurity Radar by @wavestoneFR & Bpifrance, placed in the Securing AI space under AI agent security.
A year ago, most of the conversation was about using AI to move faster. Now, as teams hand real work to autonomous agents, a second question shows up right behind it: how do you keep those agents governed, auditable, and under control? That shift from "can it act" to "can we trust how it acts" is exactly the ground we're building on.
The radar tracked 234 French cyber startups this year, worth a scroll if you want a sense of where the ecosystem is moving. Thanks to the Wavestone and Bpifrance teams for the mapping.
Native connectors or MCP servers; now you choose how your AI agents connect in MindFlow.
Broad coverage and granular control, or reach into your internal tools. Either way, enterprise-grade security is built in. 🔒
Demo: youtu.be/CJY6tSaW2go?si=VMbQ…
If your CRM automation stops at lead sync, the issue is not marketing. It’s integration depth.
Contacts, campaigns, messages, deals, customer data. They live across multiple systems.
When integrations are shallow, automation stops at data transfer.
A contact is created. Then humans manage the rest.
CRM only scales when updates, segmentation, messaging triggers, deal stages, and customer lifecycle actions are fully accessible through APIs.
Full read and write coverage. Not CSV imports. Not one-way sync.
@ActiveCampaign · @messagebird · Lead jet · @HubSpot · @salesforce
Our Head of Engineering, Gregory Roubinet, put a swarm of AI agents to the test. He built a simple agent that sends commands to a server and returns results, then connected it to a "supervisor" agent tasked with acting as a pentester.
First attempt: it refused. No authorization, no action, exactly the kind of guardrail you want.
So he added a human-in-the-loop step: one Slack message, one click to approve, and the flow picks up right where it left off. From there, the agent got to work, running nmap scans, checking services, and flagging an exposed MySQL instance in real time.
From there, adding more capability was just as simple: a couple of extra tools (Slack updates, Jira ticketing), one line in the prompt each time.
It's a good look at what agent swarms can actually do in practice: composable, tool-connected agents that get real work done, with a human still holding the risky calls.
Full video here: youtu.be/BjQfvEa_rqg?si=uL99…
Not every operation fits neatly into a deterministic workflow. That's where agent chaining and swarming come in.
Multiple agents, each with a specific role, work together to handle tasks that are too complex, too dynamic, or too context-dependent to model step by step.
And it all stays within enterprise governance standards. Every action is logged. Every agent's capabilities are scoped down to the individual API call if needed.
This clip is from a webinar we hosted with GLIMPS, Check it out here: youtu.be/oM93NKDmmCI?si=aFrC…
If your compliance workflows still depend on screenshots and manual evidence collection, the issue is not policy. It’s integration depth.
Access controls, privileged accounts, audit logs, identity provisioning, security searches. They live in different systems.
When integrations are shallow, automation stops at reporting.
You can document controls. You can’t enforce or validate them in real time.
Compliance only scales when user lifecycle actions, privilege management, audit trails, and security checks are fully accessible through APIs.
Full read and write coverage. Not exports. Not point-in-time snapshots.
@CyberArk Application Access Manager · @Imperva Audit Trail · @ironclad_inc SCIM · @Trellix MVISION Searches · @Veriff