MyDecisive
by MyDecisive
Open-source, OpenTelemetry-native telemetry control plane that acts on incidents instead of just charting them
MyDecisive is an AI DevOps platform for enterprise IT operations built on SmartHub, an open-source Kubernetes- and OpenTelemetry-native control plane that filters and acts on telemetry inside your own environment. Its commercial Octant suite adds anomaly detection with automated runbooks, cloud auto-scaling and security log optimisation for teams whose observability bill is growing faster than their reliability.
MyDecisive is an AI DevOps platform that treats production telemetry as something to act on rather than something to store and be billed for. Its open-source foundation, SmartHub, published as MyDecisive/mdai-hub under AGPL-3.0, is a Kubernetes- and OpenTelemetry-native control plane for telemetry pipelines and OTel collectors that runs inside the customer's own cloud or data centre, intercepting logs, metrics and traces at the edge instead of shipping the full firehose to a third-party observability SaaS. On top of that sits Octant, the commercial suite, which packages capabilities as modular applets: error management, self-trained anomaly detection with automated runbook execution, cloud auto-scaling, security log optimisation and additional governance controls. Practical features include intelligent log-stream filtering, intelligent trace sampling, PII redaction performed at the source before data leaves the environment, and IFTTT-style automated responses intended to cut mean time to resolution. The design point is economic as much as technical: observability spend scales with data volume, so filtering and acting on telemetry at the edge is pitched as a way to stop paying a vendor to ingest noise. The company was founded in 2023 by Ari Zilka, previously the founder of Terracotta, which sold to Software AG, an architect of the first version of Walmart.com and a technology lead at Hortonworks. It exited stealth on 15 July 2026 with a $12 million Series A led by Copper Sky Capital, formerly AZ-VC. MyDecisive says it is running in production across financial services, retail, telecommunications and healthcare, and claims roughly 90 percent lower production operating costs and single-day time to value, though both figures are vendor-reported.
The SRE or platform engineering leader whose Datadog or Splunk bill is scaling with AI-workload telemetry volume, and who wants to filter and act on that data inside their own Kubernetes cluster rather than pay to ingest it first.
Telemetry is filtered, redacted and acted on at the edge inside your own environment, so ingest volume and the bill attached to it fall while automated runbooks handle the incidents.
At a Glance
- Category
- Infrastructure & Cloud
- Pricing
- Freemium, Contact for pricing
- Target Market
- CTOs, VPs of Engineering, SRE Leaders, Platform Engineers, DevOps Engineers, CISOs
- Deployment
- Open-source, Self-hosted, Hybrid, Edge-first
- Founded
- 2023
- Headquarters
- San Francisco, United States
- Customers
- No public customer count; the vendor names Intelli Platforms as a partner and cites production deployments in financial services, retail, telecommunications and healthcare
Key Features
- ✓SmartHub open-source control plane
Kubernetes- and OpenTelemetry-native control plane for telemetry pipelines and OTel collectors, published under AGPL-3.0 and installed by Helm chart.
- ✓Edge telemetry processing
Processes logs, metrics and traces inside the customer's own cloud or data centre rather than shipping everything to a third-party dashboard.
- ✓Octant applets
Modular commercial programs covering error management, anomaly detection, cloud auto-scaling and security log optimisation, addable in hours rather than quarters.
- ✓Self-trained anomaly detection with automated runbooks
Learns normal behaviour and executes runbooks automatically when it deviates, aiming to resolve incidents before a human is paged.
- ✓PII redaction at source
Strips sensitive fields from telemetry before it leaves the environment, which is what makes third-party log destinations viable under data-handling rules.
- ✓Intelligent log stream and trace sampling
Filters routine telemetry and samples traces adaptively so ingest volume, and the bill attached to it, drops materially.
- ✓IFTTT-style automated workflows
Trigger-and-response automations wired to live telemetry streams, intended to cut mean time to resolution on known failure patterns.
Capabilities
Use Cases
- •Reducing an observability bill without losing signal
Filter routine logs and sample traces adaptively at the edge so only meaningful telemetry reaches the paid backend, cutting ingest volume.
- •Keeping AI workloads stable in production
Detect anomalies in services running AI workloads and execute runbooks or auto-scale cloud capacity before an outage reaches users.
- •Meeting data-handling rules on telemetry
Redact PII from logs and traces inside the customer environment so regulated data never reaches an external observability vendor.
- •Escaping proprietary observability agents
Standardise collection on OpenTelemetry with an open-source control plane, keeping the option to change analysis backends later.
- •Optimising security log pipelines
Route and reduce security telemetry before it hits a SIEM, where per-gigabyte ingest pricing makes unfiltered forwarding expensive.
Ideal For
Best For
- ✓Cutting observability ingest bills by filtering and sampling telemetry before it reaches a third-party SaaS platform
- ✓Keeping regulated data in-region by redacting PII at the source rather than after it has left the environment
- ✓Automating first-response incident handling with self-trained anomaly detection and runbooks instead of paging a human first
- ✓Standardising on OpenTelemetry without committing to a single observability vendor's proprietary agent
- ✓Kubernetes-heavy estates in financial services, telecommunications, retail and healthcare where telemetry volume and data-handling rules collide
Not Ideal For
- ✗Teams without Kubernetes in production. SmartHub is a Kubernetes-native control plane deployed by Helm chart, so there is no meaningful path without a cluster
- ✗Organisations whose legal or procurement policy blocks AGPL-3.0 code, which is a common restriction and applies to the entire open-source foundation
- ✗Buyers who want a full observability platform with dashboards, APM and log search. MyDecisive controls and acts on the pipeline; it does not replace the backend you send data to
- ✗Risk-averse buyers who need a large reference base, since the company exited stealth in July 2026 with one named partner and no published customer logos
Deployment
Market & Ratings
No public customer count; the vendor names Intelli Platforms as a partner and cites production deployments in financial services, retail, telecommunications and healthcare
Market Analysis
Pros
- ✓Attacks a real and growing cost line: observability spend scales with telemetry volume, and AI workloads are increasing that volume faster than headcount
- ✓OpenTelemetry-native with an open-source control plane, so there is no proprietary agent to rip out and the analysis backend stays replaceable
- ✓Processing and PII redaction happen inside the customer's environment, which is a genuine compliance advantage over shipping raw telemetry to a SaaS vendor
- ✓Founder Ari Zilka has a relevant track record, having founded Terracotta through to its Software AG acquisition and led technology at Hortonworks
Cons
- ✗The open-source claim is thin in practice: the mdai-hub repository shows roughly 3 stars and no forks, so there is a licence but effectively no community, contributors or third-party validation
- ✗AGPL-3.0 is a hard blocker at many enterprises, and it covers the entire foundation layer rather than an optional component
- ✗The headline claims, nearly 90 percent lower operating costs and single-day time to value, are published with no methodology, no named customer benchmark and no independent verification
- ✗No named production customers at all. The only external name in the launch materials is a partner, Intelli Platforms, not a reference account
- ✗No G2, Capterra, TrustRadius or PeerSpot listing and no Hacker News or Reddit discussion could be found, so there is no independent practitioner evidence to weigh
- ✗The commercial product's name collides with VMware's well-known and now-archived Octant Kubernetes dashboard, which makes independent research on it unusually hard
- ✗No published pricing, no metering unit and no security certifications disclosed, all of which a regulated buyer will need before procurement
Pricing
SmartHub (Open Source)
$0
- ✓AGPL-3.0 licence
- ✓Kubernetes- and OpenTelemetry-native control plane
- ✓Helm chart install
- ✓Telemetry pipeline and OTel collector management
- ✓Community Slack support
Octant (Commercial Suite)
Contact for pricing
- ✓Error management applet
- ✓Self-trained anomaly detection with automated runbooks
- ✓Cloud auto-scaling
- ✓Security log optimisation
- ✓Governance and security controls
SmartHub is free to self-host under AGPL-3.0, which matters commercially: that copyleft licence is blocked outright by some enterprise legal teams and obliges source disclosure if you build a derived service on it. Octant, the commercial suite, publishes no list price and no metering unit, so cost has to be established through a sales conversation. The savings case rests on the vendor's claimed 90 percent reduction in production operating costs, a headline figure published without methodology or an independently verified customer benchmark.
Security & Compliance
Connect
Sources
This page was written from 6 sources, 5 on domains other than mydecisive.ai.
- 1.siliconangle.com — ai devops startup mydecisive launches 12m open source smarth
- 2.github.com — mdai hub
- 3.prnewswire.com — mydecisive exits stealth with 12m funding to redefine enterp
- 4.citybiz.co — mydecisive emerges from stealth with 12 million to scale ai
- 5.analyticsinsight.net — mydecisive secures 12m seed funding launches smarthub open s
- 6.mydecisive.ai — mydecisive.aivendor
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
SkyPilot
Run and scale AI workloads across Kubernetes, Slurm and 25+ clouds from one interface
General Compute
ASIC-based inference cloud built for agents, where latency compounds
Cisco Cloud Control
One console where humans and AI agents run and defend the whole Cisco estate
Vector Core Compute
Disaggregated enterprise inference cloud running CPUs, GPUs and RDUs in one pipeline