Exact matches still found
Hybrid search finds part numbers, error codes and clauses.
- Home
- Agentic AI
- RAG systems
RAG for enterprise data
RAG systems over your internal enterprise data.
Your organization already holds the answers, spread across documentation, tickets, contracts and databases nobody can search properly. We build retrieval systems that answer questions from that material with citations, respecting who is allowed to see what.
- Hybrid search
- Permission-aware
- Cited answers
What you get
Answers people can verify.
Permissions respected
Filtering at query time against your existing access model.
Every claim cited
Answers link to the source passage.
Measured, not guessed
An evaluation set built from real user questions.
The problem
The demo answered three questions beautifully.
Naive RAG works impressively on a curated sample and degrades badly on a real corpus. Pure vector search misses exact matches like part numbers, error codes and contract clauses, because semantic similarity is the wrong tool for identifiers. Chunking splits tables and procedures mid-thought, so retrieved context is incoherent. Permissions get ignored, so the assistant cheerfully surfaces the salary review to whoever asks. And with no evaluation set, nobody can tell whether last week's prompt change helped or quietly broke retrieval.
What we build
What's included.
- Ingestion across documents, wikis, tickets, email and databases, with incremental re-indexing
- Hybrid retrieval combining vector similarity with keyword search, so exact identifiers are still findable
- Structure-aware chunking that keeps tables, procedures and clauses intact
- Permission-aware retrieval enforced at query time against your existing access model
- Citation-backed answers linking to the source passage, so users can verify rather than trust
- An evaluation set and retrieval metrics so every change is measured, plus honest refusal when the corpus lacks the answer
How we work
How we run it.
Step 1
Collect real questions from the people who will use it, and build the evaluation set from those
Step 2
Ingest a bounded, high-value corpus first rather than everything at once
Step 3
Measure retrieval quality before tuning generation, since most failures are retrieval failures
Step 4
Pilot with a real user group, review the questions it fails, and expand deliberately
What we use
Tools, platforms and methods.
- Anthropic Claude
- OpenAI
- pgvector
- Elasticsearch
- LangGraph
- Python
- Postgres
Related work
Projects we've delivered.
FAQ
Common questions.
Why RAG instead of fine-tuning a model on our data?
They solve different problems. RAG changes what the model can see at answer time and is the right answer for factual, changing information: update a document and the next answer reflects it, with a citation. Fine-tuning changes how a model behaves, such as tone, format or a specialized task. Fine-tuning on facts is expensive, goes stale immediately and cannot cite sources.
How do you handle permissions and confidential material?
Filtering happens at retrieval time against your existing access model, so a user's query can only ever match documents they are already entitled to read. Permissions are never handled by asking the model to withhold information, because that is not a security boundary.
What stops it inventing answers?
Three things together. Answers are constrained to retrieved context, every claim carries a citation the user can open, and the system is built to say it does not know when retrieval returns nothing relevant. That last behavior is the one teams skip, and the one that determines whether people keep trusting it.
How do you measure whether it is any good?
With a labeled evaluation set built from real user questions. We track retrieval metrics, such as whether the correct passage appeared in the top results, alongside answer correctness. Without this, tuning is guesswork and regressions go unnoticed for weeks.
Related
Related services.
Report presenter agent
A meeting agent that presents your performance report on the call and captures follow-ups.
Learn moreFine-tuning for ad copy
On-brand ad copy at volume, with platform limits and claims checked in code.
Learn moreAI-led companies
RAG, agents, evaluation harnesses and LLM red teaming for production AI.
Learn moreScope a RAG pilot.
Tell us which questions your team struggles to answer today and where the answers live.