Persistent search and collection
Run repeatable searches, connect external sources, preserve results, and make collection status visible.
For investigation and research teams
Software for investigation and research teams
I build focused tools that bring collection, structured findings, search, relationship analysis, and exports into one dependable workflow.
Problems I can work on
When the workflow is getting in the way
Journalism, OSINT, due-diligence, and research teams working with many sources, documents, entities, and relationships.
- Evidence is split across tools
- Links, documents, notes, spreadsheets, and external search results are difficult to review as one investigation.
- Sources change or disappear
- A useful source needs a repeatable collection and archiving workflow, not a browser bookmark someone has to remember.
- Entities are hard to connect
- Names, companies, events, documents, and locations need structure before a team can follow relationships confidently.
- AI output cannot be trusted blindly
- Generated findings need schemas, validation, source context, and a review step before they become operational data.
- Handoffs create duplicate work
- Researchers repeatedly reformat the same material for colleagues, reports, exports, and downstream systems.
- The tool is not built for daily use
- A promising prototype still needs authentication, permissions, failure states, observability, and production operations.
What the engagement can deliver
A useful system, not another layer of process
Structured investigation records
Turn documents and findings into validated entities, relationships, notes, and reviewable source material.
Graph and data exploration
Give researchers practical search, filters, record views, and relationship exploration without hiding the underlying evidence.
Production delivery
Cover the web application, APIs, data model, integrations, exports, tests, deployment, and ongoing improvement.
Relevant experience
A production investigation data platform
I have worked across architecture, backend, frontend, and production delivery for a system that supports investigations from external data collection through structured findings, relationship analysis, and export.
Read the anonymous case study- Persistent search and external integrations
- Schema-validated LLM-assisted structuring
- PostgreSQL records and Neo4j graph exploration
- Authentication, exports, and operational workflows
When this work fits
- You already understand the research workflow and need an engineer to turn it into dependable software.
- You have a working product that needs a new data source, workflow, interface, or production hardening.
- You want one senior engineer who can work across the interface, API, data model, integrations, and deployment.
Questions before we start
If the situation is not covered here, send a short description of the workflow, product, or delivery gap.
- Can you join an existing research product?
- Yes. I can own a defined product area, integration, data workflow, or frontend while working with an existing product and engineering team.
- Do you build with AI?
- When it improves a real workflow. I treat model output as untrusted input and add schemas, validation, source context, review states, and failure handling.
- Can the work stay anonymous?
- Yes. The public case study can describe the problem, contribution, and technical work without naming the organization or exposing sensitive data.
- Do you only take complete rebuilds?
- No. A focused improvement to a working system is often a better engagement than replacing the product.
Tell me about the work
Send a project brief or details about a role. I work in English and Russian and usually reply within one or two business days.