Skip to content
Book a session

Work

Selected work

Production AI platforms I've designed and built for enterprise teams, used by 500+ business users across healthcare and retail. Client names are withheld.

Enterprise platform

Enterprise Agent Platform

Multi-agent workflow orchestration

A platform for building, running and governing multi-agent workflows across the enterprise. I architected its backend orchestration layer.

  • FastAPI
  • LangGraph
  • MCP
  • Function calling
  • Guardrails
  • PII redaction
  • STT / TTS
  • FinOps

Highlights

  • Graph-based execution with FastAPI and LangGraph: sequential and parallel agent tasks, conditional branching and state transitions.
  • MCP servers as a standard tool layer, so agents discover and invoke enterprise tools, data sources and APIs through function calling.
  • Guardrails and PII detection and redaction on inputs, tool calls and outputs, enforcing content policy and data privacy.
  • Agent FinOps: per-agent token metering, cost attribution by model and provider, budgets with alerts, and actual-vs-forecast spend dashboards.
  • Real-time voice chat with agents through speech-to-text and text-to-speech pipelines, alongside text channels.

Related services: AI Agents & Automation, AI Strategy & Audits

Biopharmaceutical client

Healthcare Conversational Intelligence

Ask-your-data analytics over Snowflake

LLM-driven conversational workflows, built with LangChain and FastAPI, that turn business questions into contextual analytical requests over enterprise healthcare data.

<2sP95 latency under high concurrency
−40%analytics backlog
  • LangChain
  • FastAPI
  • Snowflake
  • NL-to-SQL

Highlights

  • Translates natural-language questions into contextual analytical requests over enterprise data in Snowflake.
  • Built for real-time analytical workloads under high concurrency.
  • Accelerated self-service insights for business stakeholders.

Related services: Ask-Your-Data Analytics

Retail client

AI Insight Portal

Decision intelligence platform

Monitors business KPIs across enterprise retail datasets and surfaces what changed, how much it matters and why.

  • KPI signals
  • Anomaly detection
  • Variance analysis
  • LLM insights

Highlights

  • Configurable signal framework for KPI monitoring, threshold evaluation, severity scoring and signal execution.
  • Automated anomaly and variance analysis that detects unusual KPI movements and triggers downstream insight generation.

Related services: Ask-Your-Data Analytics

Want something like this for your team?

Tell me what you're trying to build. In a free 30-minute call we'll work out whether AI is the right tool, what it would take, and whether I'm the right person to build it.