Case Study // Gridwise

Building a Secure AI Assistant for Gig Worker Insights

How Turon AI built a natural language data assistant with strict PII guardrails, ReAct agent architecture, and retrieval-augmented SQL generation, deployed as an MVP to 100 gig workers.

Domain
Gig Economy Platform
Scope
MVP / ~100 Workers
Stack
LangGraph + BigQuery + Vertex AI
Delivered By
Turon AI

Secure Assistant

Guarded Query Flow

MVP

User Query

Natural language

01

Pre-Query Filter

Intent + PII check

02

ReAct Agent

LangGraph + few-shot

03

BigQuery

SQL execution

04

Post-Response Filter

PII scrub + compliance

05

Safe Response

Chainlit UI

06

Natural Language

No SQL or dashboards required

Strict Isolation

PII blocked across the pipeline

Enable natural language queries over worker data while strictly preventing access to other workers' PII.

Traditional dashboards were too rigid. Open-ended AI systems risked exposing sensitive data. The challenge was to combine the flexibility of a natural language interface with strong data governance and safety guarantees without compromise on either.

The Challenge

Flexibility and safety cannot be a trade-off

Natural Language Interface

Workers needed to ask questions about personal performance, earnings, and regional market trends without using dashboards or SQL.

Strict Data Isolation

The system had to prevent access to other workers' data and block personally identifiable information at every layer.

Safety Guarantees

The assistant needed to avoid hallucination, over-sharing, and data leakage even when users asked open-ended questions.

The Approach

Designed from first principles

Turon AI designed the MVP from the ground up, focusing on three core principles that shaped every architectural decision. Early in development, multiple approaches were explored for text-to-SQL generation, including tools like Vanna.

After evaluation, the team standardised on a ReAct-style agent architecture using LangGraph, which provided better control over reasoning, tool usage, and extensibility than simpler alternatives.

01

Controlled Data Access via SQL

All data access goes through structured querying: explicit, auditable, and schema-constrained.

02

Layered Safety Mechanisms

Guardrails are applied before and after generation, not only at one checkpoint.

03

Built for Extensibility

The architecture can evolve with new agents and capabilities without a redesign.

The Solution

Eight components, one coherent system

User Query

Natural language

Pre-Query Filter

Intent + PII check

ReAct Agent

LangGraph + few-shot

BigQuery

SQL execution

Post-Response Filter

PII scrub + compliance

Safe Response

Chainlit UI

01

Natural Language to SQL via ReAct Agents

A LangGraph-powered ReAct agent interprets queries, generates and executes SQL against BigQuery, and returns structured, user-friendly responses.

02

Multi-Layer Safety and Guardrails

Strict data boundaries are enforced through pre- and post-processing filters applied at both ends of every query.

03

Retrieval-Augmented Few-Shot Learning

Semantic retrieval of few-shot examples via vector search on Vertex AI improved accuracy and reduced hallucinations.

04

Multi-Agent Extensibility

An experimental supervisor layer supports future multi-domain expansion without architectural redesign.

05

Multi-LLM Flexibility

The platform can work with OpenAI, Gemini, and other providers, with dynamic switching and no core code changes.

06

Insight Generation from Structured Data

The system generated proactive automated insights grounded entirely in real data, moving beyond reactive Q&A.

07

Interactive UI with Chainlit

An interactive chat interface made the system accessible during the MVP phase and fast to iterate.

08

Observability and Iteration with LangSmith

LangSmith instrumentation gave visibility into agent decisions and supported rapid iteration from real usage.

Deployment & Iteration

Launched to 100 real workers

The system was launched as an MVP to a controlled group of approximately 100 gig workers. This allowed the team to observe real-world usage patterns, identify edge cases and failure modes, and continuously refine the system based on actual behaviour.

Every observation from the field fed directly back into prompt refinements, filter adjustments, and agent behaviour updates.

01

Observe real-world usage patterns

Monitor how workers naturally phrase questions and what they most want to know.

02

Identify edge cases and failure modes

Catch query types that expose gaps in SQL generation or filter logic.

03

Refine prompts, filters, and agent behaviour

Targeted updates based on LangSmith traces and real interaction data.

Outcome

Powerful, private, and production-ready

01 / Access

Natural Language Insights

Workers could access personal performance and regional market data without dashboards or SQL knowledge.

02 / Safety

Strict Privacy Boundaries

The system maintained strict data privacy and access boundaries with zero PII leakage and zero cross-user data exposure across all tested queries.

03 / Accuracy

Improved SQL Accuracy

Retrieval-augmented prompting improved query accuracy and reduced hallucinations across varied and complex questions.

04 / Architecture

Scalable Multi-Agent Foundation

New domains and capabilities can be added without redesigning the core system.

05 / Flexibility

Beyond Fixed Dashboards

A flexible query interface reduced dependency on rigid dashboards by adapting to worker questions.

The result is an AI assistant that does not just answer questions, but does so safely, accurately, and in alignment with real-world constraints.

Key Takeaway

Combining AI capability with system design discipline, grounding generation in structured data, enforcing layered safety, and building for extensibility, is what makes AI assistants both powerful and trustworthy.

AI capability and system design discipline

Getting natural language querying right is not just a model problem. It is an architecture problem, a safety problem, and an iteration problem all at once.

By grounding generation in structured data through SQL, enforcing safety through layered guardrails, and designing for extensibility from day one, Turon AI delivered a system that is both powerful and controlled.

The MVP deployment to 100 real workers created the tight feedback loop needed to refine the system rapidly. The result is a foundation that can grow with new agents, new domains, and new capabilities.