Dev Tools & APIs

AWS Rekognition is a cloud-based image and video analysis service from Amazon Web Services that uses machine learning to identify objects, people, text, scenes, and activities in visual content. It also provides facial analysis, face comparison, and content moderation capabilities — all accessible through a straightforward API. Businesses use AWS Rekognition for a wide range of applications: automating image moderation on user-generated content platforms, extracting text from scanned documents, verifying identities through facial comparison, analysing security footage, and categorising product images for e-commerce catalogues. At Osher, we integrate AWS Rekognition into n8n workflows that process visual content automatically. For example, when a new image is uploaded to your system, a workflow can analyse it with Rekognition, extract relevant metadata, moderate the content, and route the results to your database or notification system. We have built similar AI-powered document processing pipelines — see our medical document classification case study for an example. Learn more about our approach on our AI agent development page. If your team is manually reviewing images, categorising visual content, or extracting information from photos and scanned documents, AWS Rekognition combined with automated workflows can handle that work faster and more consistently.

AMQP (Advanced Message Queuing Protocol) is an open-standard messaging protocol used for passing messages between applications, services, and systems. The AMQP Trigger node in n8n listens for messages on an AMQP queue and starts a workflow whenever a new message arrives — making it the bridge between your message broker and your automation layer. Development teams, DevOps engineers, and system architects use AMQP-based message brokers like RabbitMQ and ActiveMQ to build reliable, decoupled systems. Messages queue up when services are busy, nothing gets lost during traffic spikes, and different parts of the system can scale independently. Osher uses the AMQP Trigger to connect message-driven architectures to n8n automation workflows. When your application publishes a message — a new order, a data update, a processing request — the AMQP Trigger picks it up and routes it through whatever workflow logic you need: data transformation, API calls to external systems, database updates, or notification delivery. Our system integration services specialise in connecting message brokers to the rest of your tech stack. If your systems use message queues and you need to connect them to business workflows, get in touch to discuss your architecture.

Sentry.io is an application monitoring and error tracking platform used by development teams to detect, triage, and fix software issues in real time. It captures errors, exceptions, and performance problems across web, mobile, and backend applications, giving engineers the context they need to resolve issues fast. Software teams, DevOps engineers, and product managers rely on Sentry to keep applications stable. Instead of finding out about crashes from angry users, Sentry surfaces problems the moment they happen — with full stack traces, breadcrumbs, and environment data attached. Osher integrates Sentry into broader operational workflows using n8n. When Sentry detects a critical error, we can automatically create tickets in your project management tool, alert the right team member on Slack or Teams, and log the incident for compliance reporting. Our AI agent development work also uses Sentry data to build intelligent triage systems that categorise and prioritise errors without manual intervention. If your dev team is drowning in unstructured error alerts, talk to us about connecting Sentry to the rest of your workflow.

QuestDB is a high-performance time-series database built for speed. It uses a column-oriented storage engine and supports SQL queries, making it accessible to anyone who already knows SQL while delivering query performance that outpaces traditional relational databases on time-series workloads by orders of magnitude. It is designed for scenarios where millions of rows need to be ingested per second and queried with sub-second response times. Businesses use QuestDB for IoT data ingestion, real-time application monitoring, financial market data analysis, and operational analytics. Its PostgreSQL wire protocol compatibility means it works with existing BI tools like Grafana, Metabase, and Tableau without custom connectors. Data can be ingested via the InfluxDB Line Protocol, CSV import, or REST API, giving teams flexibility in how they feed data in. Osher builds automated data processing pipelines on top of QuestDB for clients who need fast analytics on high-volume time-series data. We handle the deployment, schema design, ingestion pipeline setup, and dashboard creation — connecting QuestDB to your data sources and visualisation tools so insights are available in real time. If your current database is struggling with time-series query performance or you are looking for a purpose-built solution for high-frequency data, reach out to discuss QuestDB.

uProc is a data enrichment and processing API that cleans, validates, and enhances business data in real time. It offers a library of processors for tasks like email verification, phone number formatting, company data lookup, text normalisation, and data type conversion — all accessible through a single API. Instead of juggling multiple data quality tools, teams can run dozens of data operations through one platform. For businesses dealing with messy CRM data, unverified lead lists, or inconsistent formatting across systems, uProc solves a real problem. It can validate email addresses before they enter your marketing platform, standardise phone numbers across international formats, enrich company records with firmographic data, and clean text fields that have accumulated garbage over time. Osher integrates uProc into automated data processing workflows so data gets cleaned and enriched as it moves between systems — not as a manual batch job someone remembers to run quarterly. We connect uProc to your CRM, marketing tools, and databases so every record meets your quality standards before it reaches a human. If bad data is costing your team time and trust, talk to us about building a data quality pipeline with uProc at its core.

Google Cloud Natural Language is a machine learning API that analyses text to extract sentiment, entities, syntax, and content categories. The n8n integration lets you feed text from any source — emails, support tickets, reviews, documents — into the Natural Language API and use the analysis results to drive automated decisions in your workflows. Businesses use it to automatically classify incoming support tickets by sentiment and topic, extract company names and product mentions from customer feedback, analyse survey responses at scale, and categorise documents by content type. Instead of reading and sorting text manually, the API processes it in seconds and returns structured data your workflows can act on. At Osher, we build text analysis pipelines that connect Google Cloud Natural Language to your business systems. Whether you need to route negative customer reviews to your support team, extract entity data from contracts, or classify documents for compliance, we design workflows that turn unstructured text into actionable data. We have done similar work for clients in healthcare and insurance — see how we handled medical document classification using AI. If your team is manually reading and categorising text data, contact us to discuss how natural language processing can automate that work.

Grafana is an open-source observability platform that turns metrics, logs, and traces from your systems into visual dashboards and alerts. If you’re running automated workflows, APIs, databases, or infrastructure, Grafana shows you what’s actually happening — and more importantly, alerts you when something goes wrong before your customers notice. Most businesses don’t monitor their automations. Workflows fail silently, APIs degrade without warning, and nobody knows there’s a problem until a customer complains. Grafana connected to your business automation stack changes that — giving you real-time visibility into everything that’s running. How we use Grafana in automation projects: Monitoring workflow execution rates, success rates, and processing times Tracking API response times and error rates across integrated services Alerting teams when queue depths grow, indicating processing bottlenecks Visualising data pipeline throughput and identifying slow stages Dashboarding business KPIs pulled from multiple data sources in real time Grafana connects to virtually any data source — Prometheus, InfluxDB, PostgreSQL, MySQL, Elasticsearch, CloudWatch, and dozens more. This means you can monitor your data processing pipelines, AI agents, and system integrations from a single pane of glass. Our consulting team builds Grafana dashboards and alerting for organisations that need visibility into their automated operations. When you’re running dozens of workflows, APIs, and integrations, you need to know they’re all working — Grafana is how you get that confidence without manually checking every system.

Google Cloud Realtime Database (Firebase RTDB) stores and syncs data across connected clients in milliseconds. It’s the database behind real-time applications — live dashboards, chat systems, collaborative tools, and IoT data streams. When data changes in Firebase RTDB, every connected client sees the update instantly without polling or manual refreshes. The challenge with real-time data is doing something useful with it beyond the app itself. When a new record lands in Firebase, you might need to update a CRM, trigger an alert, process the data through an AI model, or sync it to a warehouse. That’s where workflow automation connected to Firebase RTDB becomes valuable. Automations we build with Firebase Realtime Database: Processing IoT sensor data as it arrives and triggering alerts on threshold breaches Syncing real-time application data to analytics warehouses and reporting tools Feeding user activity data from mobile apps into CRM and marketing systems Triggering AI-powered analysis on new data entries as they’re written Maintaining backup copies of critical real-time data in secondary storage Firebase RTDB is particularly useful for automated data processing scenarios where timing matters. Sensor readings, user actions, and live data feeds can trigger processing workflows within seconds of the data being written, rather than waiting for batch jobs to run. Our consulting team integrates Firebase RTDB into broader system integration architectures and AI agent workflows. Whether you’re building real-time dashboards, processing IoT data, or syncing mobile app activity to your back-office systems, we connect Firebase to the tools that turn raw data into business action.