AI in Software

TaxWorld (Ireland/UK) built Ezylia, an AI tax assistant on CustomGPT.ai no-code platform, handling 2,000+ queries/day at 97.5% resolution across 189k queries, doubling ARR to ~EUR 1m in 24 months.

Anthropic's CI team built Claude Tag, an AI agent that serves as the first responder for CI/CD test failures, typically publishing its first analysis within 15 minutes and naming the root cause within 4 minutes in fastest cases. The agent, wired into Datadog, Grafana, PagerDuty, GitHub, Kubernetes, and Slack via MCP Connectors, has replaced hour-long manual investigations with three-minute fix verifications across internal CI infrastructure.

Indonesian healthtech Halodoc has rolled out AI-automated code review across its full-stack engineering organization, layered on top of its existing Jenkins Global Library CI/CD pipeline. The deployment addresses consistent-feedback challenges as Merge Request volume grew across engineering repositories.

Salesforce's MuleSoft unit has built a trust-bar system that runs against AI-generated code to prevent merges when reliability checks fail. The system responds to the velocity problem that agentic development creates, where AI-produced code reaches merge boundaries faster than legacy manual review cycles can absorb.

Robo-advisor Wealthfront built its own in-house AI code review harness over multiple years, after an engineer pasted a misbehaving branch diff into GPT-3.5 and caught a missed SQL filter instantly. The custom harness overcomes limits of general-purpose tools like Claude Code and raises trust in AI-augmented review by combining diff-level bug catching with exploration via sub-agents.

ThinkDeep AI's DeepBrain multi-agent platform has been deployed across the French Ministry of Economy and Finance and the French Ministry of Defense, saving €2M at scale across 10,000 Ministry of Finance employees, cutting document retrieval from two days to two minutes.

Paris-based multiplayer AI platform Dust deployed Claude via Anthropic’s API to power its no-code AI agent infrastructure, saving roughly $10,000 per day on model spend (18-19% reduction) after optimizing prompt caching. The deployment raised autonomous tool-calling depth from 4-5 to 8+ steps per agent run with no engineering changes and now serves more than 300,000 agents deployed across customers including Datadog, Vanta, and 1Password.

Austin-based SaaS company SupportFlow deployed the ECOA AI customer-support platform and cut first response time 73% (14.2 hours → 3.8 hours), raised CSAT from 4.1 to 4.6, and nearly halved annualized agent turnover (32% → 14%) within 90 days by automating 58% of incoming tickets end-to-end.