Paitho

Devtools &
Infrastructure.

For companies selling to platform teams, SREs, and backend engineers. Built for the post-Heroku generation: the buyers who read your changelog before they read your homepage.

PAIN SIGNALS
87
PITCH ANGLES
12
AVG REPLY RATE
11.8%
OPERATORS LIVE
34
Tuned by
Ex-Stripe GTM · Ex-Datadog SDR lead · Two devtools founders who shipped the cold email themselves.
Devtools & Infrastructure
Sandbox runs on synthetic leads. No data leaves your browser.

The full vocabulary your buyers actually use — when they're not on a sales call.

Detection runs across hiring posts, GitHub repos, public incident retros, conference talks, and changelogs. Frequency counts are rolling 30-day mention volume across the operator pool.

Infrastructure
kubernetes_sprawl▲ 247 heroku_migration94 multi_cloud_cost76 vpc_peering_pain52 terraform_drift47
Reliability
observability_gap▲ 198 stale_runbooks142 on_call_burnout131 slo_violations88 deploy_freeze71 ci_flake_rate49
Cost & spend
datadog_bill_shock63 snowflake_credits58 egress_surprise44 finops_role_open38
Hiring & team
platform_team_ramp▲ 118 staff_eng_gap54 secrets_leaked68 dx_metrics_open39 first_sre_hire31
▲ trending up · last 30 days. Full taxonomy is exposed in the sandbox; weights and half-lives are versioned per release.

Every angle pinned to a pain signal. Every reply rate published.

Angle Triggered by Sample subject line Reply rate
runbook_automation stale_runbooks · on_call_burnout the 3am rollback you wrote up last month 17.2%
oncall_rotation_pain on_call_burnout · staff_eng_gap your sev-2 graph from the q4 incident report 14.8%
platform_ramp_cost platform_team_ramp · finops_role_open 4 platform-eng job posts in 6 weeks — what's the plan 12.1%
post_migration_audit heroku_migration · terraform_drift after the heroku cutover, the second bill 10.3%
observability_gap observability_gap · slo_violations the metric you wish you'd been emitting in march 9.7%
k8s_consolidation kubernetes_sprawl · multi_cloud_cost the 14th cluster question 8.9%
deploy_freeze_winback deploy_freeze · ci_flake_rate thursdays without a deploy 8.4%
datadog_bill_intervention datadog_bill_shock · finops_role_open the line item nobody on your team knows how to lower 7.6%
first_sre_hire first_sre_hire · platform_team_ramp before you write that sre job posting 7.1%
secrets_hygiene secrets_leaked · dx_metrics_open the credential rotation nobody scheduled 6.4%
dx_metric_starter dx_metrics_open · staff_eng_gap your first dora dashboard 5.8%
generic_intro (no signal — control) quick question about your stack 1.2%
The control row exists to keep us honest. We benchmark every new angle against generic_intro on a holdout sample. If it can't beat 1.2%, it doesn't ship.

What a research-grounded draft actually reads like.

Three real drafts, redacted. Each opens with a citation — a public incident, a hiring post, a recent talk — and closes with a small ask. No deck. No "10 minutes."

Draft 01 · angle = runbook_automation · signal = stale_runbooks
From:    [REDACTED-OPERATOR]
To:      [REDACTED-PROSPECT]
Subject: the 3am rollback you wrote up last month

Maya — saw your incident retro on the February regional failover.

The bit where an SRE had to grep commit history to find the last
working runbook is exactly the fire we help platform teams put out.
Two engineers at [REDACTED-CUSTOMER] cut MTTR 41% after we replaced
their wiki runbooks with structured ones tied to alert IDs.

No deck. 12-min loom if useful.

— Jordan
Draft 02 · angle = platform_ramp_cost · signal = platform_team_ramp
From:    [REDACTED-OPERATOR]
To:      [REDACTED-PROSPECT]
Subject: 4 platform-eng job posts in 6 weeks — what's the plan

Pat — noticed you opened a fourth platform-eng req on March 11th,
and your golden-path doc on the eng blog hasn't been touched
since November.

fwiw the math usually goes: every new platform hire spends ~6
weeks reverse-engineering what the last one built. We sit in
between — codify the path so the next req ramps in 9 days, not 9
weeks. Worth a 10-min look at how [REDACTED-CUSTOMER] did it?

If timing's wrong, ignore. No follow-up cadence on this one.

— Sam
Draft 03 · angle = datadog_bill_intervention · signal = datadog_bill_shock
From:    [REDACTED-OPERATOR]
To:      [REDACTED-PROSPECT]
Subject: the line item nobody on your team knows how to lower

Hey Ravi — your finops req went up last Tuesday. The JD calls out
"observability spend optimization" twice. tl;dr: that's almost
always custom metrics cardinality, and it's almost always
fixable in a week without dropping coverage.

We mapped it for [REDACTED-CUSTOMER] last quarter — 38% off the
monthly bill, same alert fidelity. Happy to send the audit
template even if you don't talk to us.

— Iris

The qualification YAML. Versioned. Auditable. Yours to override.

devtools.scoring.v1.4.yaml expand ▾
pack: devtools_infrastructure
version: 1.4
weights:
  digitalMaturity:    0.28   # public eng blog, github org, conference talks
  channelDependency:  0.16   # platform-led growth signals
  productFit:         0.22   # explicit pain match (top 3 signals)
  companySize:        0.18   # 50–800 engineers
  accessibility:      0.16   # contact route exists, not gated by sdr-only forms

tier_thresholds:
  A: 0.78
  B: 0.62
  C: 0.45

hard_disqualifiers:
  - engineering_headcount: < 20
  - infrastructure: bare_metal_only
  - last_funding_round_age_months: < 12
  - no_public_facing_infra: true

half_life_days: 45            # signal staleness decay
refresh_cadence_hours: 168    # weekly re-score

Override any weight per campaign. Disqualifiers are hard — if a company trips one, it's not scored, it's removed. We'd rather miss a fit than waste a domain.

How devtools buyers actually write.

Dry. Lowercase when in doubt. Code references over adjectives. Sentences sometimes start with "fwiw" or "tl;dr" because that's how Slack messages start. The signoff is a first name, not a title — "Jordan", not "Jordan Reyes, GTM Lead." A devtools email that begins with "Hope this finds you well" is dead on arrival. The voice is engineer-to-engineer, even when it isn't. If your draft sounds like it was written for a CMO, you've already lost.

Who's running this pack in production.

Anonymized by request. We share names on a private call.

Operator 01
Founder, observability startup
Series A. 18 engineers. Selling into 200–800-eng platform teams. Runs the pack solo, three sends a day.
Operator 02
SDR lead, infra-cost SaaS
Two-person SDR team behind a 30-person company. Replaced a $7k/mo agency with the pack and one approval queue.
Operator 03
GTM, CI/CD platform
Founder-led. Uses the pack for inbound enrichment first; outbound is the second loop. ~60 sends/week.

Versioned, like the rest of the system.

v1.4
Apr 12, 2026
Tuned runbook_automation opener variants based on 1,400 sends. Dropped two underperformers. Added finops_role_open as a feeder signal.
v1.3
Mar 02, 2026
Added datadog_bill_shock, snowflake_credits, egress_surprise. New "cost & spend" category in taxonomy.
v1.2
Feb 04, 2026
Tuned platform_ramp_cost angle. Added GitHub helm-chart count as a detection input for kubernetes_sprawl.
v1.1
Jan 08, 2026
Added 14 reliability signals after the holiday-incident wave. New angle: deploy_freeze_winback.
v1.0
Dec 01, 2025
Initial public release. 64 signals, 8 angles. Tuned by ex-Stripe and ex-Datadog GTM operators.

Adjacent packs that ship with devtools.

Try the Devtools pack live.
In a sandbox, in 2 minutes.