Skip to content
View adaptivesystemsdev's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report adaptivesystemsdev

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
adaptivesystemsdev/README.md

Adaptive Systems: deterministic AI infrastructure

We build AI products where the model reads and code decides: every fact comes from config or from the page, every decision is a check you can audit, and every number we publish was measured on the live system. Three products run in production on our own server, and we build custom AI systems for clients to the same standard. adaptivesystems.dev

Product What it does Case study · try it
Front Desk AI receptionist for service businesses: voice agent (browser or a real phone number), website chat with a call button, Telegram, grounded support answers, Google Calendar booking, lead scoring behind a deterministic gate, one inbox for every channel ai-command-center · live numbers · score a lead
Docket Invoice and document intake for Europe and the US: every value traced to the page it came from (PDFs, scans and phone photos, read twice by a vision model and OCR), arithmetic / IBAN / US routing / VAT checked in code, a risk gate for bank-detail fraud (IBAN or ACH), duplicates and PO mismatches docket-invoice-intake · live numbers · try your invoice
Dossier Applicant document verification for lenders, letting agents and AML-regulated firms: bank statements, bills, payslips and IDs read by a model; code finds edits inside the PDF (before → after), text hidden under boxes, retyped amounts, broken balances and bad MRZ check digits, and compares every name, address and date with the declaration. Checks documents, never scores people dossier-applicant-verification · live numbers · forge an applicant

All three on one page: the one-pager (what they do, how they decide, measured numbers, stack, links). Docket and Dossier also work inside Claude: connect the MCP server (OAuth, a sandbox of your own, erased after 30 minutes).

Hire us for a custom build: assistants on your data, document pipelines, voice and chat agents, decision engines, n8n automations and integrations, or a rescue of a bot that misbehaves. Scoped and quoted before work starts, weekly demos, tests and a written handover. Tell us what to build · Upwork

Live: status page · activity feed

Python · FastAPI · Pydantic AI · LangGraph · TypeScript · PostgreSQL / pgvector · Next.js · n8n · ElevenLabs · PDF forensics · OCR · Docker · self-hosted

Pinned Loading

  1. ai-command-center ai-command-center Public

    Front Desk case study (Adaptive Systems): AI receptionist with voice + website chat, grounded support, Google Calendar booking, forms and lead scoring behind a deterministic gate. Live numbers on a…

    Python 1

  2. docket-invoice-intake docket-invoice-intake Public

    Docket case study: AI invoice & document intake where the model reads and code decides — every value traced to the page, risk gate for bank-detail fraud, duplicates and PO mismatches. Live numbers.