Representative engagements. NDA-honest.
We work under NDA by default. Each engagement note describes the shape of the work — context, constraints, approach, handover — at the level the client signed off on. Short descriptors stand in for logos we cannot show yet.
6 of 6 shown
- Series B fintech10 weeks
Real-time risk triage agent
Replaced a stale rules engine with a traced, eval-gated agent loop. Manual review queue shrank; reviewer confidence rose.
- Agent loop
- Policy guardrails
- OTel tracing
- Eval harness
- Fortune 500 retailer12 weeks
Content operations platform
Replaced three offshore workflows with a typed automation pipeline owned by an internal platform team.
- Durable queues
- Typed contracts
- Design system
- Observability
- Public research lab8 weeks
Literature triage pipeline
Hybrid retrieval and rerank over a multi-million document corpus with freshness-aware caching and reproducible eval on held-out queries.
- Retrieval
- Rerank
- Eval suite
- Caching
- Venture-backed mobility6 weeks
Fleet exception co-pilot
Edge-deployed assistant for dispatchers. Tool-calling with strict policy checks against live telemetry; written handover to an internal team.
- Tool calling
- Policy
- Edge deploy
- Runbook
- Managed security platform9 weeks
Alert enrichment automation
Replaced manual triage steps with an idempotent pipeline. Typed contracts between SIEM, enrichment, and analyst UI.
- Automation
- Typed IO
- SIEM integration
- Analyst UI
- Healthcare analytics12 weeks
Internal research assistant
Scoped internal AI product with audit trail, PHI-aware retrieval, and feature flags. Rolled out to research staff in phased cohorts.
- Internal product
- Audit trail
- Feature flags
- Phased rollout
Receipts, not marketing.
Every note follows the same shape — context, constraints, approach, handover, and an explicit NDA boundary. If the engagement is still sensitive, the note stays in the private set.
Architecture diagram, decision log, eval methodology, and the measurable before / after. We do not publish vanity metrics or single-number claims.
Client names and any information that could be reverse-engineered into an identifier. Every case study goes through client review before it goes live.
Dataset shapes, model prompts, and any system detail whose disclosure would weaken the client. Scope-call conversations go deeper than published cases.