New
Search traces in plain English →
Find the bug before your customers do.
Tessellary puts your logs, traces and errors in one place, so your team sees what broke, why, and who it hit, in seconds.
500
POST /v1/orders
DURATION
1.84 s
SPANS
23
POST /v1/orders
1.84 s
auth.verify_token
112 ms
cart.load
198 ms
db.query carts
131 ms
inventory.reserve
256 ms
payments.charge
1.07 s
payments-gw.authorize
221 ms
retry payments-gw.capture
736 ms
email.enqueue
74 ms
orders.write
52 ms
ERRORS · LAST 15 MIN
USED BY 4,000+ ENGINEERING TEAMS, FROM TWO-PERSON STARTUPS TO PUBLIC COMPANIES
One place for every signal. Logs, traces, errors and alerts share one timeline, so the answer to “what broke?” is never three tabs away.
01 · PLATFORM
Everything you need to run production.
Six tools that usually live in six products. Here they share one search box, one timeline and one bill, so nobody copies an ID from one tab to another at 3 a.m.
LOGS
Logs you can read
Search a billion lines in under a second. Fields are parsed for you, so you filter by customer or order instead of writing regex.
0.8 s median search
TRACES
Traces across every service
Follow one request from the browser to the database. Each hop shows its time, its logs and the code that ran.
23 services in one view
ERRORS
Errors grouped by cause
Similar crashes fold into one issue with a count, the first release it appeared in and the users it hit.
9,400 events → 12 issues
ALERTS
Alerts that stay quiet
Alert on what customers feel, like slow checkouts or failed logins, not on every CPU spike at 3 a.m.
61% fewer pages
ASK
Answers in plain English
Ask "why did checkout fail at 14:02?" and get the trace, the log line and a likely cause back.
Answers cite their source
DASHBOARDS
Dashboards your team shares
Build a board in minutes from any query, then share one link with support, product and on-call.
Live, no refresh button
02 · LOGS
Search logs the way you search your inbox.
Type a customer, an order number or a plain question. Fields are parsed as logs arrive, so filters work on day one without a single regex.
service:checkout-api level:error customer:”cus_8f2k1q”
38 lines · 0.41 s
TIME
LEVEL
SERVICE
MESSAGE
14:02:08.859
ERROR
payment failed for order 71842: upstream 502
14:02:08.861
INFO
order 71842 marked failed, customer notified
14:02:08.371
WARN
gateway timeout after 200 ms, retry 2 of 3
14:02:08.123
WARN
gateway timeout after 200 ms, retry 1 of 3
14:02:07.902
INFO
charge started amount=84.00 EUR
14:02:07.598
INFO
inventory reserved for 3 items
14:02:07.414
INFO
token verified for cus_8f2k1q
14:02:07.311
INFO
POST /v1/orders received
03 · ALERTS
Get paged for customers, not for CPU.
Write alerts in the words your customers would use: checkouts failing, logins slow, emails not sent. Tessellary checks them every minute and stays quiet otherwise.
Checkout is failing
WHEN
error rate of POST /v1/orders
IS ABOVE
2% of requests
FOR
5 minutes
THEN
page the on-call engineer and post in #checkout
CHECKOUT ERROR RATE · LAST 60 MIN
Fired 14:03
Paged Maren H. · acknowledged in 40 s
Likely cause: payment provider timeouts after the 14:01 deploy
04 · SETUP
Running in five minutes, not five sprints.
Install one package, add your key, deploy. Tessellary picks up logs, traces and errors from the frameworks you already use.
01
Install the SDK
$
npm install @tessellary/node
02
Add your project key
$
TESSELLARY_KEY=tsl_live_••••••
03
Deploy and open the app
$
tessellary.init() // that’s it
WORKS WITH
Node.js
Python
Go
Ruby
Java
.NET
PHP
Rust
Elixir
Kubernetes
1.2B
log lines searched every hour
0.8 s
median search time on 30 days of data
99.99%
uptime over the last 12 months
4,000+
engineering teams, from 2 to 2,000 people
05 · IN THEIR WORDS
We used to learn about outages from app-store reviews. Now the on-call engineer sees the broken trip, the log line and the release that caused it before anyone tweets.
MH
Maren Holt
Head of Platform
at
Northloop
06 · CUSTOMERS
Teams that stopped guessing.
Short stories from engineering teams who moved their logs, traces and alerts into one place, and what changed afterwards.
Parcelry
Logistics
Parcelry traced a lost-label bug across nine services in one afternoon
9
services in one trace
Kiteframe
Design software
Kiteframe halved its logging bill while keeping 30 days of history
48%
lower monitoring spend
Brightwell
Energy
Brightwell watches 12,000 smart meters without waking anyone at 3 a.m.
61%
fewer night pages
07 · PRICING
Simple pricing that grows with you.
Start free with one project. Pay per seat when your team joins, with generous data included. No surprise bills for a noisy week.
Hobby
$0
free forever
For side projects and early prototypes.
Start free
1 project
5 GB of logs per month
7-day history
Email alerts
Team
Most popular
$29
per seat / month
For product teams running real customers in production.
Start 14-day trial
Unlimited projects
100 GB of logs per month
30-day history
Traces and error grouping
Alerts to chat and phone
Scale
$79
per seat / month
For companies with compliance needs and many services.
Start 14-day trial
Everything in Team
1 TB of logs per month
13-month history
Single sign-on and audit log
Data kept in your region
Enterprise
Custom
annual contract
For large fleets that need their own terms and support.
Talk to sales
Everything in Scale
Private cloud option
Dedicated support engineer
Custom retention
99.99% uptime commitment
08 · QUESTIONS
Answers before you ask.
Still unsure? Our engineers answer every message, usually within a working hour.
Do I need to change my code?
Add one SDK line to your app. Tessellary reads the logs you already write and adds traces automatically for common frameworks.
How is pricing calculated?
Where is my data stored?
Can I send data from OpenTelemetry?
What happens to sensitive fields?
Can I cancel any time?
09 · FROM THE BLOG
Notes from people who carry the pager.
Guides on logging, tracing and on-call, plus what shipped this month.
Engineering
How we sample logs without losing the ones that matter
Keep every error and every slow request, drop the noise. Here is the rule set we ship by default.
Guides
Tracing a request from the browser to the database
A plain walkthrough of distributed tracing, with one real checkout request as the example.
Guides
Alerts that fire on what customers feel
Why we alert on failed checkouts and slow logins instead of CPU, and how to write your first rule.