Merryl DMello | AI × Automation × Anime 🔮
@merryldmello.bsky.social
I automate so I don't have to work hard on just one thing | 17+ yrs tech | 16+ AI products in 5mo | Teaching the lazy way | DMs open
Ex Company HR recruiting for a Rackspace project: struggling to hire
2 months of generic job postings, low return
2 months of generic job postings, low return
January 6, 2026 at 7:01 PM
Ex Company HR recruiting for a Rackspace project: struggling to hire
2 months of generic job postings, low return
2 months of generic job postings, low return
Tracking 10+ projects manually?
jira , github issues, task lists
scattered everywhere, impossible to track
my solution:
- automated tracker that pulls from jira API
- github repo scanner for updates
- claude
- daily memo with all project states
one system to rule them all
automation = sanity
jira , github issues, task lists
scattered everywhere, impossible to track
my solution:
- automated tracker that pulls from jira API
- github repo scanner for updates
- claude
- daily memo with all project states
one system to rule them all
automation = sanity
January 5, 2026 at 8:26 PM
Tracking 10+ projects manually?
jira , github issues, task lists
scattered everywhere, impossible to track
my solution:
- automated tracker that pulls from jira API
- github repo scanner for updates
- claude
- daily memo with all project states
one system to rule them all
automation = sanity
jira , github issues, task lists
scattered everywhere, impossible to track
my solution:
- automated tracker that pulls from jira API
- github repo scanner for updates
- claude
- daily memo with all project states
one system to rule them all
automation = sanity
your startup doesn't need AWS
"but everyone uses it"
so what?
"but it scales"
you have 50 users
but it's industry standard
industry standard=overpriced
i run multiple projects on hetzner
85% cheaper, 0 regrets
stop donating to bezos just because you are too chicken to believe in something else
"but everyone uses it"
so what?
"but it scales"
you have 50 users
but it's industry standard
industry standard=overpriced
i run multiple projects on hetzner
85% cheaper, 0 regrets
stop donating to bezos just because you are too chicken to believe in something else
January 3, 2026 at 4:09 PM
your startup doesn't need AWS
"but everyone uses it"
so what?
"but it scales"
you have 50 users
but it's industry standard
industry standard=overpriced
i run multiple projects on hetzner
85% cheaper, 0 regrets
stop donating to bezos just because you are too chicken to believe in something else
"but everyone uses it"
so what?
"but it scales"
you have 50 users
but it's industry standard
industry standard=overpriced
i run multiple projects on hetzner
85% cheaper, 0 regrets
stop donating to bezos just because you are too chicken to believe in something else
Postgres covers 95% of use cases
"But we need NoSQL for scale"
you have 1000 users
"But we need graph database"
postgres has graph extensions
"But we need .."
no you don't
One database, learn it well, ship fast
"But we need NoSQL for scale"
you have 1000 users
"But we need graph database"
postgres has graph extensions
"But we need .."
no you don't
One database, learn it well, ship fast
January 2, 2026 at 7:25 PM
Postgres covers 95% of use cases
"But we need NoSQL for scale"
you have 1000 users
"But we need graph database"
postgres has graph extensions
"But we need .."
no you don't
One database, learn it well, ship fast
"But we need NoSQL for scale"
you have 1000 users
"But we need graph database"
postgres has graph extensions
"But we need .."
no you don't
One database, learn it well, ship fast
2026: the year automation gets serious
Been building systems for 18 years
Most people automate the easy stuff
This year i'm automating the impossible stuff
Can't say more yet
But follow along
Things are about to get interesting
Wish you all a happy and prosperous new year!
Been building systems for 18 years
Most people automate the easy stuff
This year i'm automating the impossible stuff
Can't say more yet
But follow along
Things are about to get interesting
Wish you all a happy and prosperous new year!
January 1, 2026 at 7:46 PM
2026: the year automation gets serious
Been building systems for 18 years
Most people automate the easy stuff
This year i'm automating the impossible stuff
Can't say more yet
But follow along
Things are about to get interesting
Wish you all a happy and prosperous new year!
Been building systems for 18 years
Most people automate the easy stuff
This year i'm automating the impossible stuff
Can't say more yet
But follow along
Things are about to get interesting
Wish you all a happy and prosperous new year!
Spent weeks writing perfect spec documents
Clients never read them
now: 1-page summary + weekly demos
Demos > Documents
People don't read
People watch
Show don't tell!
Clients never read them
now: 1-page summary + weekly demos
Demos > Documents
People don't read
People watch
Show don't tell!
December 30, 2025 at 3:43 PM
Spent weeks writing perfect spec documents
Clients never read them
now: 1-page summary + weekly demos
Demos > Documents
People don't read
People watch
Show don't tell!
Clients never read them
now: 1-page summary + weekly demos
Demos > Documents
People don't read
People watch
Show don't tell!
10 years experience ≠ 10 years of learning
most people repeat year 1 ten times
real growth: new problems, new domains, new constraints
if you're comfortable you're not growing
seek discomfort systematically
most people repeat year 1 ten times
real growth: new problems, new domains, new constraints
if you're comfortable you're not growing
seek discomfort systematically
December 27, 2025 at 2:55 PM
10 years experience ≠ 10 years of learning
most people repeat year 1 ten times
real growth: new problems, new domains, new constraints
if you're comfortable you're not growing
seek discomfort systematically
most people repeat year 1 ten times
real growth: new problems, new domains, new constraints
if you're comfortable you're not growing
seek discomfort systematically
Everything has an API now
your calendar
your email
your bank
your todo list
this is the golden age of automation
if it doesn't have an API yet it will soon
learn to connect APIs = print money
your calendar
your email
your bank
your todo list
this is the golden age of automation
if it doesn't have an API yet it will soon
learn to connect APIs = print money
December 26, 2025 at 6:13 PM
Everything has an API now
your calendar
your email
your bank
your todo list
this is the golden age of automation
if it doesn't have an API yet it will soon
learn to connect APIs = print money
your calendar
your email
your bank
your todo list
this is the golden age of automation
if it doesn't have an API yet it will soon
learn to connect APIs = print money
automating too early = waste
automating too late = burned out
sweet spot: when pain > effort to automate
doing it manually 10 times? automate it
doing it manually twice? too soon
let pain be your signal
Merry Christmas everyone, you don’t have to automate today, just chill at home with the fam
automating too late = burned out
sweet spot: when pain > effort to automate
doing it manually 10 times? automate it
doing it manually twice? too soon
let pain be your signal
Merry Christmas everyone, you don’t have to automate today, just chill at home with the fam
December 25, 2025 at 8:33 PM
automating too early = waste
automating too late = burned out
sweet spot: when pain > effort to automate
doing it manually 10 times? automate it
doing it manually twice? too soon
let pain be your signal
Merry Christmas everyone, you don’t have to automate today, just chill at home with the fam
automating too late = burned out
sweet spot: when pain > effort to automate
doing it manually 10 times? automate it
doing it manually twice? too soon
let pain be your signal
Merry Christmas everyone, you don’t have to automate today, just chill at home with the fam
"We're a team of 50"
So what?
I've seen 5-person teams outship 50-person teams
Team size is a vanity metric
Output per person matters
Revenue per person matters
Headcount is just cost
P.S. : 5 person teams automate the s**t out of their processes.
So what?
I've seen 5-person teams outship 50-person teams
Team size is a vanity metric
Output per person matters
Revenue per person matters
Headcount is just cost
P.S. : 5 person teams automate the s**t out of their processes.
December 23, 2025 at 1:56 PM
"We're a team of 50"
So what?
I've seen 5-person teams outship 50-person teams
Team size is a vanity metric
Output per person matters
Revenue per person matters
Headcount is just cost
P.S. : 5 person teams automate the s**t out of their processes.
So what?
I've seen 5-person teams outship 50-person teams
Team size is a vanity metric
Output per person matters
Revenue per person matters
Headcount is just cost
P.S. : 5 person teams automate the s**t out of their processes.
Automation handles 95% of cases perfectly
Humans break everything with creativity
"what if i put emojis in the email subject?"
"what if i upload a .zip instead of .pdf?"
"What if I put some commas in csvs?"
Edge cases are why automation is hard
Plan for human stupidity (including your own)
Humans break everything with creativity
"what if i put emojis in the email subject?"
"what if i upload a .zip instead of .pdf?"
"What if I put some commas in csvs?"
Edge cases are why automation is hard
Plan for human stupidity (including your own)
December 21, 2025 at 11:04 AM
Automation handles 95% of cases perfectly
Humans break everything with creativity
"what if i put emojis in the email subject?"
"what if i upload a .zip instead of .pdf?"
"What if I put some commas in csvs?"
Edge cases are why automation is hard
Plan for human stupidity (including your own)
Humans break everything with creativity
"what if i put emojis in the email subject?"
"what if i upload a .zip instead of .pdf?"
"What if I put some commas in csvs?"
Edge cases are why automation is hard
Plan for human stupidity (including your own)
Side projects taught me more than my job ever did
Not because jobs are bad
Because side projects let you fail cheap
Shipping > perfecting
Learning > earning
Building > planning
The ROI isn't money
It's compounding skills
Not because jobs are bad
Because side projects let you fail cheap
Shipping > perfecting
Learning > earning
Building > planning
The ROI isn't money
It's compounding skills
December 17, 2025 at 7:51 PM
Side projects taught me more than my job ever did
Not because jobs are bad
Because side projects let you fail cheap
Shipping > perfecting
Learning > earning
Building > planning
The ROI isn't money
It's compounding skills
Not because jobs are bad
Because side projects let you fail cheap
Shipping > perfecting
Learning > earning
Building > planning
The ROI isn't money
It's compounding skills
Spent years chasing impressive titles
"senior engineer" "lead developer" "principal"
None of it mattered
What mattered: could i solve expensive problems?
Titles get you interviews
Problem solving gets you paid
"senior engineer" "lead developer" "principal"
None of it mattered
What mattered: could i solve expensive problems?
Titles get you interviews
Problem solving gets you paid
December 16, 2025 at 11:44 AM
Spent years chasing impressive titles
"senior engineer" "lead developer" "principal"
None of it mattered
What mattered: could i solve expensive problems?
Titles get you interviews
Problem solving gets you paid
"senior engineer" "lead developer" "principal"
None of it mattered
What mattered: could i solve expensive problems?
Titles get you interviews
Problem solving gets you paid
"señor engineer" doesn't mean writing more code
it means writing less code that does more
automating the boring parts
architecting for maintainability
teaching others to solve problems
if you're still just grinding out features you're stuck
it means writing less code that does more
automating the boring parts
architecting for maintainability
teaching others to solve problems
if you're still just grinding out features you're stuck
December 15, 2025 at 1:16 PM
"señor engineer" doesn't mean writing more code
it means writing less code that does more
automating the boring parts
architecting for maintainability
teaching others to solve problems
if you're still just grinding out features you're stuck
it means writing less code that does more
automating the boring parts
architecting for maintainability
teaching others to solve problems
if you're still just grinding out features you're stuck
The best automations make you obsolete
At HP i automated myself out of a 3-month project
They almost hired me full time because of it and it's still being used after almost 11 years
Make yourself so efficient you're either promoted or redundant
Both beat staying stuck
At HP i automated myself out of a 3-month project
They almost hired me full time because of it and it's still being used after almost 11 years
Make yourself so efficient you're either promoted or redundant
Both beat staying stuck
December 14, 2025 at 12:30 PM
The best automations make you obsolete
At HP i automated myself out of a 3-month project
They almost hired me full time because of it and it's still being used after almost 11 years
Make yourself so efficient you're either promoted or redundant
Both beat staying stuck
At HP i automated myself out of a 3-month project
They almost hired me full time because of it and it's still being used after almost 11 years
Make yourself so efficient you're either promoted or redundant
Both beat staying stuck
Claude code now runs agents async
Game changer for automation
Spin up multiple subagents in background to explore your codebase and your documents
You keep working. Zero interruption.
Subagents finish → wake up main agent → report results, you are happy
This is what i mean by "lazy engineering"
Game changer for automation
Spin up multiple subagents in background to explore your codebase and your documents
You keep working. Zero interruption.
Subagents finish → wake up main agent → report results, you are happy
This is what i mean by "lazy engineering"
December 11, 2025 at 9:28 AM
Claude code now runs agents async
Game changer for automation
Spin up multiple subagents in background to explore your codebase and your documents
You keep working. Zero interruption.
Subagents finish → wake up main agent → report results, you are happy
This is what i mean by "lazy engineering"
Game changer for automation
Spin up multiple subagents in background to explore your codebase and your documents
You keep working. Zero interruption.
Subagents finish → wake up main agent → report results, you are happy
This is what i mean by "lazy engineering"
Nobody talks about automation maintenance
The script that saves you 2 hours/day?
Breaks every 6 months when APIs change
Real cost isn't building it
It's keeping it running
Build for maintainability with grace and not just functionality for self satisfaction.
The script that saves you 2 hours/day?
Breaks every 6 months when APIs change
Real cost isn't building it
It's keeping it running
Build for maintainability with grace and not just functionality for self satisfaction.
December 9, 2025 at 9:41 AM
Nobody talks about automation maintenance
The script that saves you 2 hours/day?
Breaks every 6 months when APIs change
Real cost isn't building it
It's keeping it running
Build for maintainability with grace and not just functionality for self satisfaction.
The script that saves you 2 hours/day?
Breaks every 6 months when APIs change
Real cost isn't building it
It's keeping it running
Build for maintainability with grace and not just functionality for self satisfaction.
The biggest automation mistake: automating broken processes
If a process sucks manually, automating it just makes it suck more and faster
Fix the process first
Then automate the fixed version
Automate efficiency not inefficiency
Remember: Garbage In Garbage Out
If a process sucks manually, automating it just makes it suck more and faster
Fix the process first
Then automate the fixed version
Automate efficiency not inefficiency
Remember: Garbage In Garbage Out
December 8, 2025 at 1:21 PM
The biggest automation mistake: automating broken processes
If a process sucks manually, automating it just makes it suck more and faster
Fix the process first
Then automate the fixed version
Automate efficiency not inefficiency
Remember: Garbage In Garbage Out
If a process sucks manually, automating it just makes it suck more and faster
Fix the process first
Then automate the fixed version
Automate efficiency not inefficiency
Remember: Garbage In Garbage Out
17 years in tech
Biggest lesson: problems are the same everywhere
different companies but chaos follows everywhere
The folks who win are the ones who build systems to handle the chaos
not the ones who just work harder
Biggest lesson: problems are the same everywhere
different companies but chaos follows everywhere
The folks who win are the ones who build systems to handle the chaos
not the ones who just work harder
December 6, 2025 at 8:16 PM
17 years in tech
Biggest lesson: problems are the same everywhere
different companies but chaos follows everywhere
The folks who win are the ones who build systems to handle the chaos
not the ones who just work harder
Biggest lesson: problems are the same everywhere
different companies but chaos follows everywhere
The folks who win are the ones who build systems to handle the chaos
not the ones who just work harder
17 years in tech
Biggest lesson: problems are the same everywhere
different companies but chaos follows everywhere
The folks who win are the ones who build systems to handle the chaos
not the ones who just work harder
Biggest lesson: problems are the same everywhere
different companies but chaos follows everywhere
The folks who win are the ones who build systems to handle the chaos
not the ones who just work harder
December 6, 2025 at 1:18 PM
17 years in tech
Biggest lesson: problems are the same everywhere
different companies but chaos follows everywhere
The folks who win are the ones who build systems to handle the chaos
not the ones who just work harder
Biggest lesson: problems are the same everywhere
different companies but chaos follows everywhere
The folks who win are the ones who build systems to handle the chaos
not the ones who just work harder
Everyone wants to build fancy AI agents
meanwhile businesses are paying for:
- automated data entry
- email classification
- document processing
- report generation
Boring AI makes money
Fancy AI makes X threads.
Don't get me wrong, I like fancy agents too, as long as they translate into returns.
meanwhile businesses are paying for:
- automated data entry
- email classification
- document processing
- report generation
Boring AI makes money
Fancy AI makes X threads.
Don't get me wrong, I like fancy agents too, as long as they translate into returns.
December 5, 2025 at 9:29 AM
Everyone wants to build fancy AI agents
meanwhile businesses are paying for:
- automated data entry
- email classification
- document processing
- report generation
Boring AI makes money
Fancy AI makes X threads.
Don't get me wrong, I like fancy agents too, as long as they translate into returns.
meanwhile businesses are paying for:
- automated data entry
- email classification
- document processing
- report generation
Boring AI makes money
Fancy AI makes X threads.
Don't get me wrong, I like fancy agents too, as long as they translate into returns.
I've heard this time and time again:
if your business breaks when you take a weekend off
you don't have a business
you have a job you created for yourself
automate until weekends are actually weekends
now that also goes for everything that you build at your job as well. Let that sink in.
if your business breaks when you take a weekend off
you don't have a business
you have a job you created for yourself
automate until weekends are actually weekends
now that also goes for everything that you build at your job as well. Let that sink in.
December 5, 2025 at 6:39 AM
I've heard this time and time again:
if your business breaks when you take a weekend off
you don't have a business
you have a job you created for yourself
automate until weekends are actually weekends
now that also goes for everything that you build at your job as well. Let that sink in.
if your business breaks when you take a weekend off
you don't have a business
you have a job you created for yourself
automate until weekends are actually weekends
now that also goes for everything that you build at your job as well. Let that sink in.
I measure success by how little i do manually
Right now: 3 manual tasks per week
Used to be: 50
That's not lazy but me automating my workflows.
Where are you guys at in your automation.
P.S.: Not talking about n8n
Right now: 3 manual tasks per week
Used to be: 50
That's not lazy but me automating my workflows.
Where are you guys at in your automation.
P.S.: Not talking about n8n
December 3, 2025 at 8:32 PM
I measure success by how little i do manually
Right now: 3 manual tasks per week
Used to be: 50
That's not lazy but me automating my workflows.
Where are you guys at in your automation.
P.S.: Not talking about n8n
Right now: 3 manual tasks per week
Used to be: 50
That's not lazy but me automating my workflows.
Where are you guys at in your automation.
P.S.: Not talking about n8n
2017
HR couldn't hire for 2 months.
Manual screening wasn't working.
I built an automation:
- Targeted candidate profiles
- Skill filtering
- Screening questions
- Simple ranking
0 → 20 hires in 2 months.
They called it magic.
It was just automation.
HR couldn't hire for 2 months.
Manual screening wasn't working.
I built an automation:
- Targeted candidate profiles
- Skill filtering
- Screening questions
- Simple ranking
0 → 20 hires in 2 months.
They called it magic.
It was just automation.
December 3, 2025 at 10:27 AM
2017
HR couldn't hire for 2 months.
Manual screening wasn't working.
I built an automation:
- Targeted candidate profiles
- Skill filtering
- Screening questions
- Simple ranking
0 → 20 hires in 2 months.
They called it magic.
It was just automation.
HR couldn't hire for 2 months.
Manual screening wasn't working.
I built an automation:
- Targeted candidate profiles
- Skill filtering
- Screening questions
- Simple ranking
0 → 20 hires in 2 months.
They called it magic.
It was just automation.
In 2018, my manager said: "Complete 3 years of manual work in 3 months."
I said: "Give me 2 months to automate it. If I fail, I'll do overtime to finish manually."
Delivered in 1 month and 1 week.
That's when I learned: the laziest solution is usually the smartest one.
I said: "Give me 2 months to automate it. If I fail, I'll do overtime to finish manually."
Delivered in 1 month and 1 week.
That's when I learned: the laziest solution is usually the smartest one.
December 2, 2025 at 8:22 PM
In 2018, my manager said: "Complete 3 years of manual work in 3 months."
I said: "Give me 2 months to automate it. If I fail, I'll do overtime to finish manually."
Delivered in 1 month and 1 week.
That's when I learned: the laziest solution is usually the smartest one.
I said: "Give me 2 months to automate it. If I fail, I'll do overtime to finish manually."
Delivered in 1 month and 1 week.
That's when I learned: the laziest solution is usually the smartest one.