Autonomous Database Reliability
For database teams with bigger problems to solve

An autonomous DBRE that works with your team, not around it.

SIXTA is a tireless database reliability engineer that joins your team — starting with PostgreSQL and MySQL. It investigates incidents in minutes, not hours, lives in your incident and change workflows, and gets better the longer it works with you.

# db-incidents SIXTA active
SX
SIXTA DBRE 03:47 AM
Latency spike on prod-pg-primary-01 — checkout queries on analytics_events jumped from 40ms to 2.4s in the last 20 minutes.
Root Cause Analysis
Cause: Query plan regression — planner switched to sequential scan after last night's ANALYZE
Why: analytics_events grew by 12M rows from batch import. Updated statistics now estimate seq scan is cheaper than the index path
Impact: 2.4s avg → 60× slower · checkout page timeouts rising
Checked & ruled out: connection pool, lock contention, disk I/O — all normal
SX
SIXTA DBRE 03:48 AM
Immediate mitigation applied: reset planner statistics for analytics_events to restore the previous index path. Queries already recovering — down to ~80ms and dropping.
Recommended fix (requires change approval)
Proposal: Covering index on (event_type, created_at, user_id) — planner will prefer it regardless of row estimates
Expected: ~35ms restored · resistant to future stats drift
Risk: Low — CREATE INDEX CONCURRENTLY, no locks, rollback = drop index
Status: Submitted to change management · awaiting approval
MR
Maria R. 03:49 AM
Nice catch. I wouldn't have connected the CPU spike to the ANALYZE run. Stats reset is a smart interim fix — I'll get the index into tomorrow's change window.
👍 3
From a Design Partner

"Sixta helped me to spot a vacuum to prevent wraparound before static alerts. What can I say besides it works when I need and hides when I focused on other tasks."

JS
Jefferson Santos
Database Reliability Architect, AB InBev
Built by the team behind VividCortex and Percona · Backed by Baron Schwartz

Direct to your databases. Insights where you work. Everything inside your walls.
Databases
PostgreSQL
MySQL
Observability
Datadog
Prometheus
Percona PMM
AWS CloudWatch
Azure Monitor

+ extensible to new sources

SIXTA
DBRE
Delivers To
Slack
MS Teams

+ webhooks & custom

Your team's tools
SIXTA's structured output — root cause, affected queries, recommended fix — is designed to feed your existing incident workflow, whether that's ServiceNow, PagerDuty, Jira, or an internal system.
Cloud-managed
AWS RDS, Aurora, Azure Database, Cloud SQL — SIXTA connects directly.
Self-managed
On-prem, private cloud, sovereign — SIXTA deploys alongside your databases, inside your walls.
Internal DBaaS
Your platform team runs the database layer as a service. SIXTA integrates into your internal DBaaS — the reliability layer your application teams never see.
Beyond the database
SIXTA correlates database symptoms with infrastructure signals — disk I/O, CPU queues, memory pressure, OS-level errors — so your team can diagnose issues that span the stack, not just the database layer.
Deploys inside your infrastructure · Your data never leaves your environment · Read-only database access

The database didn't cause the outage. But it knew about it 40 minutes before your team did.

Every incident leaves a trail in the database layer long before it becomes a page. Latency drifts. Lock contention builds. Query plans shift. The evidence is there — but your monitoring tools treat it as background noise until something breaks.

46 min
Typical manual investigation time
per P1 incident
→ 6 min
With SIXTA on the team
root cause + remediation plan

Your SREs need to know if it's the database — in minutes, not after an hour of triage. Your DBAs need to see what changed before the escalation hits. Right now, both are staring at the same dashboards and seeing nothing useful.

SIXTA reads what the database already knows and delivers the answer to your team while the incident is still forming.

See how the investigation works →

This is your 3am right now.
03:12  PagerDuty fires: checkout latency > 2s on prod-pg-primary-01
03:14  On-call engineer opens Datadog. 14 dashboards. Nothing obvious.
03:22  Checks slow query log. Finds 3 suspects. Runs EXPLAIN on each.
03:31  Correlates with deploy log. Nothing deployed since yesterday.
03:38  Checks pg_stat_statements. Finds the plan regression after ANALYZE.
03:44  Writes the covering index. Tests in staging. Deploys to prod.
03:58  Queries back to normal. Slacks the team. Goes back to sleep.
46 minutes. $210K in downstream impact. And your best engineer just lost a night's sleep.
With SIXTA
03:12  SIXTA detects latency anomaly. Begins investigation.
03:14  Traces to query plan regression after nightly ANALYZE on analytics_events.
03:15  Rules out connection pool, lock contention, disk I/O. All normal.
03:16  Recommends covering index. Posts full RCA to #db-incidents.
03:17  Maria approves. SIXTA builds index concurrently. Zero downtime.
03:18  Queries restored to ~35ms. Monitoring for 24 hours.
6 minutes. Your engineer slept through it.

That 46 minutes happens every time. Same investigation, same manual steps, different night. SIXTA turns it into 6 minutes — and your team sleeps through it.

"It works when I need and hides when I focused on other tasks."

Jefferson Santos · Database Reliability Architect, AB InBev

Dealing with something like this? Tell us about it


Where the savings come from

SIXTA doesn't just find problems — it gives your team the capability to resolve them in minutes, prevents outages before they escalate, and turns every incident into evidence you can measure.

LOWER MTTR

Minutes, not hours

Every minute saved is thousands recovered

Every minute of database downtime costs thousands. SIXTA delivers root cause analysis in minutes — cutting mean time to resolution and the cost of every incident. At $4,537 per minute, even a modest MTTR reduction pays for itself many times over.

EXPAND CAPABILITY

Your SREs resolve database incidents without waiting for a DBA

Autonomy for the on-call, not just efficiency for the org chart

Today, every database incident ends the same way: the SRE pages a DBA and waits. SIXTA gives your on-call engineers the diagnostic depth to resolve database issues themselves — root cause, affected queries, and a recommended fix, delivered in minutes. The DBA team gets pulled in for genuinely complex problems, not routine triage.

PREVENT INCIDENTS

Detect, don't react

The cheapest incident is the one that never happens

Continuous baseline monitoring detects regressions and anomalies before they become outages. SIXTA spots the query plan regression on Tuesday, so your team isn't firefighting at 3am on Saturday.

Your DBAs get strategic. Your SREs get autonomous.
SIXTA handles the reactive database work — so your DBAs focus on architecture and your SREs resolve incidents without waiting on a DBA.
For DBAs & DBREs

Stop doing the same investigation for the third time this month. SIXTA handles the repetitive diagnostic and tuning work. You approve every change and focus on architecture, migrations, and strategy.

For SREs & Platform Teams

Fewer pages. Faster triage. When an incident fires, SIXTA has already traced it to the database layer — or ruled it out. Your on-call gets an answer, not another alert to investigate.

Without SIXTA

  • Your on-call spends 46 minutes per incident investigating manually
  • SREs escalate to the DBA team and wait. The same investigation happens again next week.
  • One senior DBA holds all the knowledge — and they're burned out.
  • Monitoring tools show 14 dashboards of symptoms. You still find the root cause by hand.
  • Your senior DBREs are expensive and stretched thin. They spend most of their time on investigations that follow the same pattern every time.

With SIXTA

  • Root cause in 6 minutes, delivered to Slack with full evidence
  • SREs can resolve database incidents directly — no DBA escalation required.
  • Every investigation documented and reusable. The knowledge compounds.
  • One answer, not 14 dashboards. The specific query, table, and fix.
  • Your team's coverage scales without adding headcount

SIXTA never acts on a database it hasn't earned the right to act on.

Autonomous database operations isn't a leap of faith. It's a progression. SIXTA starts in read-only mode, surfacing insights you can validate. After it proves itself, you expand its permissions — on your timeline, on your terms.

WEEKS 1–2

Observe

Read-only access. SIXTA connects, maps topology, learns query patterns, and builds a baseline. Nothing is modified.

WEEKS 3–6

Recommend

Root cause analysis and specific fixes delivered to Slack. Your team reviews every recommendation before acting.

WHEN YOU'RE READY

Act

Autonomous actions for risk-assessed operations. Each action type individually gated with full audit trail.


Built by database operators,
for database operators

The founding team built database monitoring together at Percona, VividCortex, and SolarWinds — then reunited to build what comes next. Backed by Baron Schwartz (founder of VividCortex), Heavybit, Encoded Ventures, and Percona.

Ewen Fortune, Founder and CEO of SIXTA

Ewen Fortune

Founder & CEO

25 years in database operations. Built monitoring at VividCortex (→ SolarWinds). Led database product at enterprise scale.

Andrew Morgan, Co-founder and CTO of SIXTA

Andrew Morgan

Co-founder & CTO

Engineering leadership across RKG, VividCortex, and SolarWinds. Built and operated multi-shard MySQL systems handling hundreds of millions of rows.

Lori Gagnon, Head of Go-to-Market at SIXTA

Lori Gagnon

Head of Go-to-Market

6+ years in database and infrastructure software across VividCortex and SolarWinds. Translating complex technical products into enterprise revenue.

Advisors

Nitay Joffe — Founder/CTO ActionIQ · Facebook Data Infra · HBase Core Contributor
Andy Ellicott — 20 years launching database companies · Vertica, VoltDB, Cloudant/IBM, DBOS
Christian Bilien — Head of Data at BNP Paribas · Global Head of DBRE at Société Générale

Our investors also backed PagerDuty, LaunchDarkly, Citus Data, Crunchy Data, and Snyk



Your CISO will approve this
SIXTA deploys inside your VPC, connects read-only to your databases, and delivers insights to your Slack or Teams. No external SaaS pipeline. No credentials leaving your walls. Ever.
Deploys inside your infrastructure
Your data never leaves your network
Your data trains your DBRE
Never used for other customers' models
Full audit trail
Every action logged, every decision explained
Learn more about our security architecture →

Questions about how SIXTA deploys in your environment? hello@sixta.ai

Ready to add a DBRE
to your team?

SIXTA is onboarded like a person. It learns your environment, earns your trust, and takes level-1 and level-2 DBRE work off your plate. Every action is evidence-backed and reviewable.

Meet Your New DBRE

Prefer email? hello@sixta.ai

Your next DBRE is ready to start

Meet Your New DBRE