Available for software engineering roles - backend systems, production reliability, cloud deployment, and AI-assisted workflows.

Connect
Privacy Infrastructureprototype

Scrapper

Local PII scrubber that acts as a security guard in front of AI workflows, grounded in TDPI paper research (March).

Case file

Main outcome

Gives businesses an always-on privacy guard for AI-touched data.

Secondary signal

Lets teams adopt AI without routing raw customer data through external APIs.

Evidence count

3 proof points

01

Problem

Businesses want AI in their workflows but cannot send raw customer data to third-party models. PII leakage is a hard blocker, not a nice-to-have.

02

Solution

Built a local Dockerised blackbox that scrubs inbound data before it reaches a model and checks outbound responses before they reach the user, with deterministic rules independent of any specific model.

03

Role

Privacy infrastructure builder · Dockerised guard-rail designer

Impact

  • Gives businesses an always-on privacy guard for AI-touched data.
  • Lets teams adopt AI without routing raw customer data through external APIs.
  • Behaviour stays deterministic and auditable regardless of the model behind it.

Technical Highlights

  • Docker-packaged local scrubber, drop-in for any AI workflow
  • Inbound request and outbound response interception
  • Deterministic PII detection rules, model-independent
  • Academic grounding from TDPI paper (March)

Proof

  • Public GitHub repository
  • TDPI paper research grounding (March)
  • Local-first, self-hosted by design

Constraints

  • Prototype stage; scrubbing rules need ongoing coverage for edge-case PII.
  • Deterministic scrubbing trades recall for predictability.

Limitations

  • Research-backed prototype, not yet a hardened enterprise DLP product.
  • Requires per-deployment tuning of rules for domain-specific identifiers.

Roadmap

  • Broaden PII rule coverage and add audit logging.
  • Add CI-tested scrub/unscrub round-trip examples.
All Case StudiesProof-of-Work LabGitHub

Next field note

Let's make the next system less fragile.

Open to software engineering roles across full-stack systems, platform and reliability work, workflow automation, and applied AI.

Write to Himanshu Lade →