Announcement: OpsGuru Signs Strategic Collaboration Agreement with Amazon Web Services Read more⟶
Announcement: OpsGuru Signs Strategic Collaboration Agreement with Amazon Web Services Read more⟶
Industry: Media & Entertainment, Digital News
Goal: Build "SaySo," an AI/ML-powered short-form news platform that automates video metadata extraction, moderation, and high-velocity discovery for the social media age.
OpsGuru Services: Cloud-Native MVP Development, AI/ML Integration, Data Pipeline Engineering, Infrastructure as Code (IaC).
AI/ML: AWS Rekognition (Visual Signals), AWS Transcribe (Speech-to-Text), AWS Comprehend (Sentiment & Topics).
Infrastructure, Backend & API: AWS AppSync (GraphQL), AWS Step Functions (Orchestration), AWS Media2Cloud, AWS CDK (Cloud Development Kit)
Data & Search: Amazon Aurora (PostgreSQL), Amazon OpenSearch Service, Amazon S3.
End-to-End Content Intelligence: Automated the ingestion of short-form video to extract transcripts, topics, and sentiment without manual intervention.
Accelerated Time-to-Market: By validating and remediating existing "partially built" services instead of restarting, the team saved months of development time.
Scalable Trust & Safety: Implemented automated moderation pipelines, allowing the platform to maintain news-grade integrity at social-media speeds.
Operational Independence: Delivered a stabilized production environment with comprehensive documentation, empowering the internal team to iterate independently.
Caliber is reimagining journalism for a digital-first culture. With a portfolio reaching over 100 million monthly users (including The News Movement, The Recount, and Capsule), the company delivers news at the speed of culture. Their latest venture, SaySo, represents a strategic evolution. Looking to move beyond social media feeds, Caliber is creating a proprietary, creator-centric platform - an "Apple News+ for creators" - where journalistic integrity is as vital as engagement.
This vision addressed a critical industry problem: the Trust Gap. To meet modern audiences on their terms, Caliber needed to deliver short-form, social media optimized video content quickly, while also ensuring that the content used was rigorously verified. This required a sophisticated backend capable of real-time analysis, tagging, and moderation. AI wouldn’t just be a feature layer, but the operating foundation that makes trustworthy, high-velocity journalism possible.
Bridging the gap between a high-fidelity iOS frontend and a complex, partially deployed AWS environment was the essential step to making this high-trust news model a reality. Caliber partnered with OpsGuru to transform this vision into a production-hardened MVP, ensuring the platform could scale while upholding news-grade editorial rigor.
Caliber faced an ambitious product vision and market pressure combined with a technical landscape that was only partially defined. With an internal team focused primarily on the iOS frontend, the road to a production-ready MVP required solving both strategic and deep technical hurdles:
Strategic & Operational Challenges: Caliber needed to demonstrate a working platform quickly to maintain market momentum and validate the product ahead of launch. Introducing complex AI features (like automated tagging and moderation) too early could have locked them into rigid, expensive architectural decisions before the core platform was stable. Caliber needed a partner to sequence these ML features so they enhanced the MVP, instead of creating delays.
The Partially Built Dilemma: Several AWS services (Aurora, AppSync, OpenSearch) were deployed but not integrated. Caliber needed to validate these existing assets without the prohibitive cost of a total rebuild.
High-Trust Automated Moderation: The MVP required an automated safety gate to process video content at scale. This involved detecting visual signals and extracting speech-to-text data to surface actionable signals for human review, ensuring editorial teams could make informed trust-and-safety decisions at speed rather than at the cost of accuracy.
Data Synchronization Hurdles: A core requirement was perfect synchronization between Amazon Aurora (the source of truth), Amazon OpenSearch (the discovery engine), and Amazon S3 (the asset store). Small misconfigurations here could lead to broken user feeds or missing search results.
Parallel Delivery Constraints: The backend work had to progress rapidly without blocking the iOS development team. This necessitated strict API contracts and a clear separation of concerns to allow the mobile experience to evolve alongside the hardened cloud infrastructure.
OpsGuru addressed these requirements through a structured three-phase engagement designed to align technical delivery with Caliber’s business milestones.
Before any material build began, OpsGuru conducted a month-long strategic assessment to define the what and the how. This phase established the reference architecture and sequenced service selection. By determining the specific mapping of features to AWS services, the team ensured the AI ambitions for automated tagging and moderation would enhance the core platform rather than delay its launch.
During a six-week high-velocity execution phase, OpsGuru embedded a virtual team to build the core backend services powering the SaySo experience. This multidisciplinary group worked alongside Caliber to implement:
Media Ingestion: AWS Media2Cloud served as the backbone for handling video assets and metadata.
Data Layer: The team utilized Amazon Aurora for application data and Amazon OpenSearch for high-speed keyword search and feed query performance.
Intelligent Media Orchestration: AWS Step Functions orchestrated an event-driven pipeline where AWS Transcribe, Rekognition, and Comprehend extracted transcripts and visual signals to ground every video in accurate metadata.
The final phase focused on validation and launch readiness. OpsGuru remediated partially deployed services and ensured the environment was production-stabilized.
Infrastructure as Code: All infrastructure was deployed using the AWS CDK to ensure repeatability across staging and production.
GraphQL API Layer: AWS AppSync was implemented as the API layer, allowing the mobile app to query multiple data sources with a single, efficient call.
“OpsGuru helped us bring structure and momentum to a complex platform build. They worked alongside our engineering team to turn partially built systems into something we could confidently move toward launch. Their ability to guide us through technical ambiguity, validate decisions early, and help us stay aligned with our launch milestones was invaluable. The result is a robust, scalable platform that will serve us well as we scale in the competitive digital news space.” — Dion Bailey, Co-Founder & CPTO, Caliber
The partnership redefined Caliber’s operational DNA, moving the SaySo platform from a vision of short-form news to a production-ready model of autonomous content intelligence.
Autonomous Content Intelligence: An automated pipeline now ingests video to extract transcripts, topics, and sentiment, freeing editorial staff to apply human judgment where it matters most: quality, context, and journalistic integrity.
Production-Ready Stability: OpsGuru synchronized the architecture (Aurora, OpenSearch, and S3) to deliver a flawless discovery engine capable of supporting millions of monthly users.
Elastic Scalability: By codifying the platform through AWS CDK (IaC), Caliber can now deploy consistent, isolated environments that grow in lockstep with its audience.
The AI Trust Gate: Automated tagging and moderation handle extraction and flagging at scale, but human review remains a deliberate part of the workflow, ensuring editorial judgment, not automation alone, determines what reaches the audience.
Ownership of the Ecosystem: By embedding a dedicated squad into Caliber’s workflow for six weeks, OpsGuru accelerated the build while simultaneously training the internal team to evolve the platform without ongoing external support.
Accelerated Innovation: The high-velocity, phased approach saved months of development time.
Solving the Trust Gap: Caliber has proven that social-speed content can maintain news-grade integrity by using AI to ground every video in verified metadata, with editorial teams using those signals to make confident, informed publishing decisions.
The Creator-Centric Blueprint: This project serves as a definitive case study for the Apple News+ for creators model, demonstrating how orchestrated cloud services can build proprietary platforms that prioritize journalistic rigor.