Your users shouldn't be the first to know.

Monitors your services, goes looking when something seems off, and pings the right engineer before your users do.

FoamAPP2:15 PM

Hey @you, P95 latency on /api/search spiked to 2,103ms. Commit a3f8e21 introduced an N+1 query in findRelatedProducts(). Batching the lookups into a single $in query fixes it.

Get Started

From scattered black boxes to one legible source.

Opens PRs across your services. Once merged, you can follow what's happening in your services back to its source.

Connect GitHubGrant read-only access to your repositories

Connect

Choose services to connect

Foam works alongside Datadog, Sentry, and whatever else you're already running. Learn how ›

Autoconfigures your first monitors while adjusting thresholds as needed.

Using your codebase and business context, sets up monitors specific to what your services do. Such as product KPIs, latency, or error rates.

Analyzing your services…

Watches for problems and tells you before your users do.

Continuously checks monitors and pings the responsible person when something needs to be looked at.

Monitoring your services

Error Rate api-gateway (5xx)0.08%

Last 1hSLO: < 0.5%42.1k req

1%0.5%0.1%0%

12:0012:1012:2012:3012:4012:5013:00

P95 Latency search-api142ms

Last 1hSLO: < 500msp50 48ms

500ms300ms100ms0ms

12:0012:1012:2012:3012:4012:5013:00

Throughput order-svc (req/s)1.2k

Last 1hPeak: 1.8kavg 1.1k

2k1.5k1k500

12:0012:1012:2012:3012:4012:5013:00

Connection Pool Saturation recommendation-svc → postgres34%

Last 1hCeiling: 80%avg 31%

80%60%40%20%

12:0012:1012:2012:3012:4012:5013:00

Ask anything. It's all in one place.

Sits on top of your highest signal data so you can ask questions such as who is affected, what else to investigate, and what to look at next.

Hey @marcus, your last deploy is causing a latency spike.

P95 latency on /api/search spiked to 2,103ms (threshold: 500ms). This is happening because your commit a3f8e21 replaced the batched findRelatedProducts() call with a per-item loop, creating an N+1 query (~48 sequential DB round-trips per request).

Suggested fix: Replace the per-item loop in src/handlers/search.ts:142 with a batched $in query.

Ask me anything...

Built for

Engineers who want to know if their code is working as intended.

Teams with fragmented observability looking to unify it.

Engineers who got pinged by a customer before their own monitors caught anything.

Teams debugging difficult production issues such as latency and silent errors.

Built for teams at

EilamStream · CTO

“Their CEO joined our channel and did not leave us until Foam was helpful.”

ShubhamLica · Founding Engineer

“We were getting forty alerts a day. Now we get three a week, and the ones we get are real.”

SteveNoto · CTO

“Before our users even get to ping us, Foam has already explained the issue and pointed us to the cause.”

AviAtlas · CTO

“I no longer lose a day checking my code changes in prod.”

SaulFounder & CTO

“Since frontend and backend apps logs become linked, I get to have the full picture.”

EilamStream · CTO

“Their CEO joined our channel and did not leave us until Foam was helpful.”

ShubhamLica · Founding Engineer

“We were getting forty alerts a day. Now we get three a week, and the ones we get are real.”

SteveNoto · CTO

“Before our users even get to ping us, Foam has already explained the issue and pointed us to the cause.”

AviAtlas · CTO

“I no longer lose a day checking my code changes in prod.”

SaulFounder & CTO

“Since frontend and backend apps logs become linked, I get to have the full picture.”

Get Started

Your users shouldn't be the first to know.

Monitors your services, goes looking when something seems off, and pings the right engineer before your users do.

FoamAPP2:15 PM

Hey @you, P95 latency on /api/search spiked to 2,103ms. Commit a3f8e21 introduced an N+1 query in findRelatedProducts(). Batching the lookups into a single $in query fixes it.