apidays Singapore 2025
2025-04-15..2025-04-16
Singapore, Singapore
[apidays Singapore 2019, apidays Singapore 2024, apidays Singapore 2025]
as the organisers put it
Singapore is a global hub for fintech, AI, and digital transformation—and the perfect place to discuss the next wave of API-driven innovation. From Open Banking to AI-driven automation, the region is setting the pace for the future.
who it's for
api architects, data engineers, machine learning engineers, platform engineers, product managers, security engineers, developer experience engineers
topics
green LLM lifecycle assessment; conversational AI agents at scale; streaming lakehouse with Kafka Flink Iceberg; UX for digital carbon reduction; AI-powered data APIs; AsyncAPI v3 and event design; multi-protocol evented API interfaces; logical data fabric for generative AI; GraphQL integration with LLMs; observability and guardrails for generative AI; API governance and automation; cross-cloud high-availability at scale
| title | author | length |
|---|---|---|
| The Quest for the Greenest LLM | Jean Philippe Ehret | 19m |
Independent audit of a major but unnamed LLM supplier compares a 3 billion and 120 billion parameter model across cradle‑to‑grave life cycle and shows the… | ||
| Building Conversational AI Agents at Scale | Thor Schaeff | 10m |
ElevenLabs shows a full pipeline that transcribes speech, routes to an LLM, hits a document RAG knowledge base and returns speech with assigned voices per… | ||
| Streaming Lakehouse with Kafka, Flink and Iceberg | Zabeer Farook | 28m |
Claimed end-to-end streaming lakehouse that ingests synthetic pageclicks with Kafka Connect data-gen, enriches and joins streams with static files using Flink… | ||
| Designing for change: how UX can help solve the climate crisis. | Julie Schiller | 11m |
Argues that UX can cut real emissions by shrinking page weight, deleting unused content, and changing defaults; gives concrete numbers like 50 grams per large… | ||
| From Data to Insights: Building AI-Powered Data APIs | Asif Waquar | 46m |
Describes an implemented claims pipeline that ingests hundreds of thousands of PDFs/images, runs OCR then an LLM for extraction and evaluation, and stores… | ||
| Beyond Bolt-On AI: A New Way to Build Software | Alan Hsiao | 32m |
Argues that the legacy OLTP/OLAP split and SaaS AI APIs force moving data to algorithms and cap automation at 5–10% automation; cites chips $626B… | ||
| AsyncAPI v3: Streamlining Event-Driven API Design | Ankit Kumar | 30m |
AsyncAPI v3 serves as spec-as-config that autodeploys infrastructure and removes doc/implementation drift; demo converts an AsyncAPI HTTP spec and a Kafka spec… | ||
| Multi-Protocol APIs: Designing a Unified Interface for EDA | Naresh Jain | 29m |
Defines a unified interface that maps channels, topics and patterns (request‑reply, pub/sub, event streaming) across Kafka and Google PubSub, with JSON… | ||
| Empowering Generative AI with Unified Data Integration: A Logical Approach. | Shanmuga Sunthar Muniandy | 27m |
Claims that poor data is the main reason GenAI projects fail, citing 66% of execs unhappy with outcomes and 72% naming data as a top barrier; enterprises… | ||
| Interacting with LLMs via GraphQL | Leonardo Losoviz | 18m |
Built and released Gato GraphQL in 2023 and demonstrates replacing the JavaScript glue layer so one single GraphQL query can fetch a post, call OpenAI Chat… | ||
| Trustworthy Generative AI: The Role of Observability and Guardrails. | Santanu Dey | 23m |
Argues that enterprise GenAI needs runtime AI guardrail as API gateway to filter malicious prompts and outputs, plus end-to-end observability that traces… | ||
| From API Intelligence to API Governance [...] | Harsha Chelle | 23m |
Analysis of ~1 billion requests across ~15,000 APIs and ~500,000 endpoints shows 85% lack rate limits, 52% accept unauthenticated calls, and only ~70% use… | ||
| Fostering API Productization in Federated API Architecture | Abhijit Dey | 26m |
Claims that federated API management enforces one centralized policy plane while giving teams autonomy, reduces central burden, boosts resiliency and… | ||
| Surviving an interconnected world with API governance | Mun Kiat Kok | 22m |
Agentic AI increases API usage and turns APIs into product channels via MCP (model context protocol, launched by Entropic) that lets agents call web APIs and… | ||
| Generative AI Landscape Building a Modern Data Strategy | Senthil Kumar | 34m |
Roadmap for generative AI centers on data quality as the foundation with 80% data quality determining model outcomes, uses GANs and VAE-style models, and… | ||
| Hybrid OnPrem + Cloud Architectures | Sagun Garg | 37m |
Claim: hybrid architectures cut long‑term AI and mission‑critical costs by moving heavy training to cloud and predictable inferencing on‑prem, with banks… | ||
| Enabling a more integrated government through APIs | Nick Loh | 20m |
APEX is a centrally managed government API platform that now hosts 33,000 APIs from about 38 publishers and handles 150M requests/month, moved from on‑prem to… | ||
| Navigating the Treacherous AI Chasm | Max Ong Zong Bao | 11m |
Maps the gap between prototype and rollout by naming timelines and teams: a 3–6 month prototype phase with a tiny cross-functional team (AI engineers, UX… | ||
| Managing 10s of thousands in TPS in a cross-cloud with high availability. | Bipin K | 21m |
Autodesk runs tens of thousands TPS through Apigee as the north gateway while backends stay in AWS, then migrates Apigee to GCP creating an extra network hop… | ||
| Enhancing Developer Productivity with UX | Faith Ang | 15m |
Apex powers 3,000 APIs across 38 agencies and handles 150 million requests a month while designers apply five concrete UX moves: prioritize clarity over… | ||
| 4 identity factors you didn't know you needed to support in your SaaS. | Daizen Ikehara | 22m |
Four concrete identity requirements for landing large customers: support multi-tenancy with per-customer branding and org objects, implement granular… | ||
| What exactly are AI Agents? | Aki Ranin | 29m |
Defines three concrete agent types — prompt agents, workflow agents, worker agents — and maps real tools to each: chain‑of‑thought loops, Manis and Copilots… | ||
| nothing matches | ||