01
Onboarding happens on a stranger's network
Provisioning is where most connected products lose customers, and you control none of the conditions.
ThinkDebug
Practical deep dives on software architecture, AI systems, and scaling digital products - drawn from the work we ship.
Devices, telemetry and companion apps designed as one product rather than three.
We build the software around connected devices: firmware-adjacent gateways and provisioning, telemetry ingestion at scale, companion apps, and the fleet tooling that lets you update and support hardware already in the field.
Connected products fail on the parts that are invisible in a demo. Provisioning a device onto a stranger's home network, updating firmware without bricking it, and supporting a unit whose owner cannot describe what is wrong are the problems that decide whether a product survives its first year.
The volume problem is different from ordinary web scale. Devices are chatty, unreliable and permanent - once shipped, they will keep sending data for years - so ingestion cost, retention policy and backwards compatibility are all decided long before you know what you actually need.
The constraints that shape every build in this sector, and the ones teams usually meet too late.
Provisioning is where most connected products lose customers, and you control none of the conditions.
Over-the-air firmware has to be resumable and reversible, because a half-applied update is a returned unit.
Once shipped, a fleet sends data for years. Ingestion cost and retention are decided before you know what you need.
Some devices will not update. The backend has to keep supporting protocol versions you would rather retire.
Diagnosing hardware through a customer's description requires telemetry designed for support, not just for analytics.
A vulnerability in shipped hardware cannot be patched by redeploying a server.
Swipe to read all 6
Each one is a system we have built before, not a capability we are willing to try.
Setup flows that survive real home networks, captive portals and impatient users.
High-volume ingestion with aggregation and retention that stays affordable.
Staged, resumable, reversible firmware rollouts with health gates.
Mobile and web apps that stay usable when the device is offline or misbehaving.
The telemetry and consoles support needs to resolve a fault without a site visit.
Certificates, key rotation and secure onboarding across the device lifecycle.
Engagements across sectors - the same team, the same standards.
Partnering with Thinkdebug to develop the Forever 52 mobile app has been a game-changer for us. They were always open to feedback, and it was clear they were as invested in our success as we were. What really sets them apart, though, is the amazing support we've received since the app launched. They’ve made sure the app continues to run smoothly and are always ready to fix any issues that come up.DMDhruv MehtaBusiness Head, Forever 52
Chosen for fit and long-term maintainability, never for novelty.
Send us the scope and we come back with an itemised estimate - no discovery fee, no obligation.
What we design and document against, from the first commit rather than the week before an audit.
Certification itself belongs to your entity and its assessor. What we provide is software built and documented so the assessment is about evidence rather than remediation.
Most of our work comes from clients we already shipped for. These are the reasons they give.
Staged rollouts, health gates, resume and rollback - because a bricked unit is a return, a refund and a review.
Captive portals, weak signal and unusual routers are where connected products lose customers, so that is where we test.
Sampling, aggregation and retention are decided up front, because a shipped fleet cannot be told to send less later.
Protocol versioning is designed in, since some devices will never update and still have to be supported.
Diagnostics are designed around the questions support will ask, not only the metrics product wants.
Companion apps degrade gracefully when hardware is offline, which is most of what users experience during a fault.
Let's start with you.
Your information stays private. We reply within 24 hours.
Still have a question we have not answered?
Ask an engineerWhat we build for wearables & iot, the work we have shipped, and what we have written about it.
