case study
The client came to us with an unstable platform and a difficult vendor transition. We secured the handover, stabilized their 2 TB production database, and shipped three AI automations that lifted the manual workload off their operators.
TECHNOLOGY STACK
REACT
The client's fleet management platform had outgrown its previous vendor relationship, and decided to hand over their platform to the OTAKOYI engineering team. This handover was the first hurdle, since the process had to be fast and safe for the client operations.
The platform underneath wasn't in shape for what came next. A technical audit surfaced critical SQL injection risk in a function called across hundreds of code paths, hardcoded API tokens to live third-party services, and permissive production configuration. These gaps had to be closed before any new build.
The client wanted to embed AI across their operations — automated trip reports, tachograph compliance tracking, an in-product co-pilot — but their 2 TB operational database had become so heavy that a single backup ran for two days. The platform required thorough optimization.
We treated this engagement as two connected projects under one roof.
The foundation project moved from inheritance to ownership. We executed a controlled vendor handover, brought security and access governance up to standard, re-architected the data layer, and added the test coverage.
With the platform stable, we shifted into our standard AI automation methodology — the approach we run for every automation engagement: an AI Operations Audit that maps and scores every workflow, a build phase delivering production-grade automations integrated with the client's existing systems.
We ran a structured technical audit across the backend, the mqtt-receptor service, and the authentication prototype — covering SQL injection surface, credential handling, access control, and transport security. The findings shaped the remediation roadmap and gave the client a documented baseline of where the platform actually stood.
The picture wasn't unusual for a platform that had grown faster than its engineering practices.
Vanguarder was an engineering challenge before it was an AI engagement. The system we inherited had to be made safe to operate, then made operable at scale, then opened up for AI. The client trusted the sequencing, and that's what let us ship three production automations on a platform that's now stable enough to keep building on.
Andrii Polishchuk
Tech Lead, OTAKOYI
The 2 TB problem was concentrated in a single table that recorded every GPS coordinate, engine reading, brake event, and sensor signal from every vehicle. That one table accounted for roughly 90% of the database's storage footprint.
The platform's reports and operational queries rarely needed deep historical data — most relied on the last 30 to 90 days.
Years of cold records were being kept in the same hot tables that served live operations, with no separation between data that needed to be fast and data that needed to be available.
With the platform stable, we moved into embedding AI into the day-to-day operations.
The goal was automating the repetitive, high-volume tasks distributed across the product's three lines: vehicle tracking, CCTV, and tachograph compliance.
A structured audit across the client's three product lines identified where AI would deliver the most value.
Scope alignment
Defined the three product lines and the operator roles to interview.
Operator interviews
Spoke with transport managers and fleet operators to map how the platform was actually used day to day, surfacing where the repetitive load lived.
Workflow inventory
Documented every candidate workflow across the three product lines as input for scoring.
Scoring
Evaluated each workflow on automation potential, impact on operator time, and integration complexity.
Prioritization
Shortlisted one priority automation per product line, ready for build.
Daily and weekly reports were assembled by hand from raw telemetry. We replaced it with an AI-generated reporting layer that synthesizes telemetry, driver behavior, and event data into a finished report, distributed to the right transport manager automatically.
Key features:
Tachograph compliance is rule-heavy, infringement-prone, and carries direct financial exposure under UK and EU driving-hours regulations. The co-pilot analyzes driver hours data continuously, flags potential infringements before they cross the line, and routes proactive alerts to the responsible transport manager.
Key features:
Transport managers spent meaningful time on ad-hoc lookups — vehicle status, behavior events, footage retrieval — work that interrupted their day without producing leverage. The co-pilot embeds a natural-language assistant directly into the platform, collapsing those lookups into seconds.
Key features:
We came to OTAKOYI in a difficult position, leaving a vendor we'd outgrown, with a platform that needed both stabilization and a real path toward AI. They executed the handover cleanly, fixed the foundations our previous partner had let drift, and then delivered the automation we'd wanted for years across all three of our product lines. Two years on, they're not a vendor we work with — they're the engineering team behind our product.
Transport Operations Director
UK Fleet Management Company
We have a proven track record of building high quality solutions for customers all over the world.
CONTACT US
The key to a successful project is a strong business idea backed by real market need, a solid tech solution, and a clear go-to-market plan.
