We’re all building an archive we’ll never finish reading
Think about what’s actually sitting on your phone right now. A decade of photos you meant to organize. Text threads with people you love. Half-finished voice memos, browsing history, health app data. Almost nobody plans for what happens to any of it after they die. That gap is exactly why a new kind of role — the digital legacy custodian — is starting to take shape across tech, law, and the funeral industry.
Somebody has to deal with what’s left behind. Accounts need closing or archiving. Families want access to photos that might otherwise vanish behind a forgotten password. Fraud spikes sharply on the accounts of people who’ve recently died. And a small but growing number of families are now asking companies to build AI chatbots or avatars trained on a dead relative’s texts, emails, or voice recordings.
A digital legacy custodian is the person who ends up handling all of this. Part data engineer, part security specialist, part ethicist, sitting at an intersection nobody quite mapped out yet.
What does a digital legacy custodian actually do?
Nobody’s settled on one job title for this yet. It shows up under different names depending on where you work — cloud companies, dedicated “digital estate” startups, funeral-tech firms, even AI labs fielding memorial requests they never expected. The actual work tends to cluster around a few things.
Designing consent-based data lifecycle systems. Someone decides, while still alive, what happens to each piece of their data after death: wiped entirely, handed to a specific person, archived, or in rarer cases turned into a memorial product.
Building access control and fraud prevention. This means verifying that whoever’s asking for a dead person’s account actually has legal standing — an executor, a next of kin with paperwork, someone with a court order. Fraud targeting deceased accounts is a real and growing problem, and it’s one of the core things a digital legacy custodian is hired to prevent.
Migrating and preserving data long-term. Formats need to stay readable in thirty years, which matters more than it sounds, since many platforms hosting this data today won’t exist by then. [INTERNAL LINK: link to your post on long-term data archival best practices]
Working with legal teams. Most laws here were written before anyone owned a “digital asset,” and they still vary by country and state. In the U.S., the Revised Uniform Fiduciary Access to Digital Assets Act is the main framework fiduciaries rely on to get legal access to a deceased person’s accounts.
Building AI memorial tools, when requested. Some families explicitly ask for this. It means wrestling with real questions about consent and grief, and whether talking to a chatbot trained on a parent’s old texts actually helps anyone.
Building deletion pipelines. For the group that wants the opposite — guaranteed, verifiable erasure instead of preservation — which is probably the larger group overall.
Why the digital legacy custodian role matters more than it sounds
It’s not the technology that’s remarkable here. It’s the scale of what’s coming.
The infrastructure gap
Every person alive with a smartphone has years of digital history piling up. Every one of them will die eventually. Most platforms were built assuming an account owner who’s alive and simply forgot a password, not one who’s permanently gone. As the first generation to grow up fully online starts aging, the gap between how much data we generate and how little infrastructure exists to handle it afterward keeps widening. Someone has to close that gap, and right now almost nobody owns the problem.
The ethical frontier
There’s also an ethical dimension nobody’s fully worked through. Building a digital stand-in for someone who died raises questions engineering teams haven’t faced before. Did the person consent to this specific use? Is it healthy for a grieving spouse or child to interact with it? Who owns the “personality” a model picked up — the family, the platform, no one? These aren’t hypothetical debates. Companies are making these calls right now, for real families, often without much precedent.
Skills every digital legacy custodian needs
Data engineering fundamentals
You’ll need real fluency in pipeline design: extracting, transforming, and archiving huge, messy datasets — photos, texts, audio, video, structured records, all mixed together. Long-term preservation is its own subfield. The formats and storage strategies that keep something readable in three decades look different from what most enterprise data engineering worries about. Because this data is unusually sensitive, solid encryption and access-control experience matters too.
Security and identity verification
This means building identity-verification flows robust enough to confirm someone’s legal right to a dead person’s data. Think wills, power of attorney, court orders. It also means knowing the fraud patterns specific to deceased accounts, since that’s a growing target. Zero-knowledge and end-to-end encryption are increasingly relevant, since more systems now let a person control posthumous access without the platform seeing the content early.
Legal and regulatory literacy
You don’t need a law degree. But you should understand digital estate law basics where you’re working, including frameworks like RUFADAA in the U.S. Privacy law gets tricky here too. GDPR and CCPA were largely written with living people in mind, so applying them to the dead gets murky fast. Being able to translate ambiguous law into concrete product decisions is a real, valuable skill.
AI and ethics, for memorial-focused roles
This means knowing how to fine-tune or prompt a model into a believable persona from a limited dataset — someone’s writing, a handful of voice recordings. It also means having a genuine grounding in grief psychology, or working closely with people who do. Badly designed memorial AI can cause real psychological harm instead of comfort. Consent has to be built into the system from the start, never defaulted to.
How to become a digital legacy custodian
Step 1: Build core data engineering skills
Start with ETL pipeline design, database architecture, and cloud storage systems like AWS S3 or Google Cloud Storage. Lean toward long-term archival thinking rather than real-time processing.
Step 2: Specialize in security and identity systems
Go deep on verification workflows, encryption architecture, and access control design. This is the backbone that separates a trustworthy system from a liability.
Step 3: Study digital estate law basics
Learn how digital assets do and don’t fit into existing inheritance law. It’ll make you far more effective at turning legal requirements into system design.
Step 4: Build a portfolio project
Design a prototype where a user can set granular rules for different categories of their data after death — delete this, archive that, transfer this to a named person — with real access-control logic behind it.
Step 5: Study existing products in the space
Research how legacy contact features from major platforms and smaller memorial-tech startups currently handle this problem. Find the gaps. [INTERNAL LINK: link to your review of digital estate platforms]
Step 6: Pair technical skill with ethical rigor
If you’re drawn to the AI memorial side, study existing discussion on AI grief technology. Be ready to push for conservative, consent-first design even when more is technically possible.
Step 7: Network outside of tech
Talk to estate lawyers, funeral directors experimenting with digital services, and grief counselors. This field is small and cross-disciplinary, and relationships here surface opportunities that don’t show up in typical job listings.
Tools of the trade
- Cloud storage and archival: AWS Glacier, Google Cloud Archive
- Identity verification platforms: borrowed from fintech and legal tech
- Encryption frameworks: end-to-end and zero-knowledge libraries
- Pipeline tools: Apache Airflow, dbt
- AI fine-tuning and voice cloning tools: used carefully, only in explicit, consent-driven contexts
Digital legacy custodian salary and job outlook
Nobody’s hiring for the exact title “digital legacy custodian” yet. This work currently lives inside broader roles: data engineer, security engineer, or trust-and-safety specialist, at companies building estate or memorial features. As the space matures and dedicated startups grow around it, expect that to change. Pay should track senior data engineering and security roles, plus a premium for the rare person who can do both the technical work and the legal or ethical thinking.
Common pitfalls to avoid
Underestimating the legal complexity. What looks like simple engineering — transfer this account to a family member — is often legally ambiguous and varies by jurisdiction. Involve legal expertise early.
Building memorial AI without genuine consent. This is the single biggest ethical risk in the field: creating an AI representation of someone who never explicitly agreed to it while alive.
Treating this as a purely technical problem. The people this work serves are often in profound grief. The best engineers combine technical rigor with real empathy for how their systems will be experienced.
Ignoring long-term format durability. Data archived today needs to remain accessible decades from now, a much longer horizon than most software teams plan for by default.
Final thoughts
The digital legacy custodian role sits at one of the most human intersections in tech: mortality, memory, and data. It’s not as glamorous-sounding as “AI researcher,” but it addresses a need every person on earth will eventually have. For engineers who want technically substantial work with real emotional and ethical weight, this is a career path worth watching as it matures from a scattered set of features into a defined profession.