AI Product Development Company
Turn an AI prototype into a real SaaS product with auth, billing, multi-tenant infrastructure, and deployment.
Explore AI Product DevelopmentAI Search Product
A Smarter Way to Search Workflow Libraries
Users needed fast access to automation workflows for both conceptual discovery and deterministic ranking. Pure semantic search was unreliable for sorting and counting. We designed a hybrid retrieval architecture that routes different intents to different systems.
Hero Outcome
10+hrs saved/week
Time reclaimed from manual searching across scattered tools
System Snapshot
The visible product result was backed by a deliberately scoped architecture and delivery plan.
Frontend
Streaming chat with rich markdown rendering, code-block copy support, and clear loading states.
Next.js 16
React 19
Semantic Retrieval
Conceptual queries hit summary and technical vector collections, merging results for relevance.
Qdrant Dual Collections
OpenAI text-embedding-3-small
Structured Retrieval
Deterministic queries route to spreadsheet backend. No vector search ambiguity for exact operations.
Google Sheets
Sorting/Ranking/Counting
4 week delivery sprint
1 engineer build team
34+ production commits
Business Impact
Executive Outcomes
10+hrs/week saved
Teams get answers in seconds instead of hunting across tools
100%answer accuracy
Every response grounded in verified source data
1interface
Replaced 5+ separate search and lookup tools
5xROI in year one
Time savings across the team repay the build cost within months
The Challenge
“Discovering and evaluating automation workflows required navigating multiple data sources. Semantic search worked well for conceptual queries like "workflows about email automation" but failed on deterministic requests like "show the 10 newest workflows" or "count workflows in this category."”
Semantic search returned irrelevant results for deterministic queries like 'show the 10 newest workflows'
Counting and ranking queries produced inconsistent answers because vector similarity cannot count or sort
Technical deep-dive requests returned hallucinated download links and incomplete code context
No way to validate whether the AI was actually calling the right tool for each query type
Users had to manually search across multiple data sources to find and evaluate workflow templates
The Transformation
What changed after we built the system
Before
Semantic search returned irrelevant results for deterministic queries like 'show the 10 newest workflows'
After
Deterministic queries route to Google Sheets for exact sorting, ranking, and counting operations
Before
Counting and ranking queries produced inconsistent answers because vector similarity cannot count or sort
After
Conceptual discovery queries use Qdrant vector search across summary and technical collections
Before
Technical deep-dive requests returned hallucinated download links and incomplete code context
After
Technical requests fetch raw workflow JSON from Google Drive with round-tripped source URLs that prevent hallucination
Before
No way to validate whether the AI was actually calling the right tool for each query type
After
100% benchmark score validates tool-calling accuracy across multiple model providers and configurations
Before
Users had to manually search across multiple data sources to find and evaluate workflow templates
After
A single chat interface searches all three retrieval systems with automatic intent-based routing
Why hybrid retrieval outperforms pure semantic search
Vector search is excellent at understanding intent. Ask 'workflows about automating customer emails' and it returns relevant results even if the exact words do not appear in the workflow titles.
But ask 'show me the 10 newest workflows sorted by date' and semantic search falls apart. There is no embedding that captures recency or count. The vector space has no concept of 'newest' or 'ten.'
Routing these queries to Google Sheets gives exact answers. A spreadsheet can sort by date and return exactly 10 rows. The AI decides which tool to use based on query intent, and the model benchmark proves it makes the right choice 100% of the time across every tested provider.
How We Built It
Technical architecture for the curious
Frontend
Streaming chat with rich markdown rendering, code-block copy support, and clear loading states.
Next.js 16React 19AI SDK 5 StreamingCustom Markdown
Semantic Retrieval
Conceptual queries hit summary and technical vector collections, merging results for relevance.
Qdrant Dual CollectionsOpenAI text-embedding-3-small
Structured Retrieval
Deterministic queries route to spreadsheet backend. No vector search ambiguity for exact operations.
Google SheetsSorting/Ranking/Counting
Technical Retrieval
Deep-dive requests fetch actual workflow JSON. Source URLs round-trip through tool outputs to prevent hallucination.
Google DriveRaw JSON InspectionSource URL Grounding
Next.js 16
React 19
AI SDK 5
Qdrant
Google Sheets
Google Drive
OpenRouter
TypeScript
Engineering Decisions
Tradeoffs we made and why
34commits
1engineer
4weeks
0test files
Three specialized retrieval tools instead of one general-purpose search
Benefit
Each query type gets the retrieval system that handles it best, with no compromises
Cost
System prompt must be carefully tuned to route queries to the correct tool every time
Strict multi-step behavior caps on tool calling
Benefit
Prevents runaway tool-calling loops that burn tokens without producing useful results
Cost
Complex queries requiring 4+ steps get truncated before reaching a complete answer
Source URL round-tripping through tool outputs
Benefit
AI can only cite URLs that actually exist in tool responses, preventing hallucinated links entirely
Cost
Tool output payloads are larger, increasing token usage per response
Qdrant over pgvector for embedding storage
Benefit
Dedicated vector database with built-in collection management, filtering, and efficient similarity search
Cost
Additional infrastructure to deploy and manage compared to an embedded PostgreSQL solution
Certain client names, proprietary workflows, screenshots, and internal assets referenced in this case study are protected under a non-disclosure agreement and have been anonymized or omitted to comply with our confidentiality obligations.
Need an AI assistant that actually gets it right?
Book a free 30-minute call. We will assess your search and retrieval needs, identify where AI keeps hallucinating, and design a system that gives accurate answers every time.
30 minutes with Apurva. Not a sales call.
Book Your Free Audit
Proof of Work
More Case Studies
B2B SaaS Platform
Business Impact
100% call coverage
Every sales call scored and coached automatically AI Sales Call Coaching Platform
Went from manual call QA to a launch-ready coaching SaaS in 8 weeks. Multi-tenant scoring, RAG playbook retrieval, and tiered billing made scaling predictable.
Next.js pgvector RAG Trigger.dev
Technical Proof
Before: 10% of calls sampled
Now: Every call reviewed
Conversational Sales Automation
Business Impact
30% more revenue
From faster engagement and automated follow-ups Conversational Sales Automation
Turned inbound conversations into qualified opportunities with automated qualification, CRM updates, bookings, call handling, and follow-ups in one system.
Conversational AI Supabase Trigger.dev
Technical Proof
Before: Leads slipped away
Now: 30% more revenue
Data Intelligence Pipeline
Business Impact
$70K+ saved/year
Replaces a full-time research analyst's output Workflow Intelligence Pipeline
Scaled workflow intelligence without budget surprises by hard-capping evaluation cost per item. Daily scraping, AI scoring, archiving, and search run in one predictable pipeline.
Trigger.dev Qdrant RAG OpenRouter
Technical Proof
Before: Manual scouting
Now: 250/day automated