Skip to content
View brohum10's full-sized avatar

Block or report brohum10

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
brohum10/README.md

Soham Jindal

Purdue Computer Engineering · AI/ML minor · Expected May 2028

Seeking Summer 2027 software engineering internships in backend, infrastructure, and applied AI.

Portfolio · LinkedIn · Email

I like projects where correctness matters after the happy path: recovery, concurrent edits, and decisions made with incomplete evidence. If you have a few minutes, these are the three repositories I would open first.

Start here

Chronicle LSM Store — A C++20 embedded key-value engine with atomic write-ahead-log batches, recovery, SSTables, and compaction. Start with the architecture and verification notes; the local benchmark includes workload and environment details.

Converge — A local-first TypeScript board with deterministic conflict resolution, offline edits, WebSocket sync, and PostgreSQL replay. The architecture notes explain the conflict model, cursor safety, and limits of the demo.

QueuePilot — A support-routing workbench with a finite human-review queue and a reproducible model comparison on the BANKING77 official test split. The evidence keeps a failed candidate release decision and its errors visible.

For another backend example, see the concurrent web crawler, including a search evaluation on human-judged queries. For an LLM workflow, see the incident copilot, including its documented misses and abstention limitation. The repository list also contains smaller learning projects.

Experience

  • ADT — AI & Automation Software Engineer: Python, data validation, and AI workflow testing.
  • L3Harris — Software Engineering Intern, ML Systems: Telemetry pipelines and anomaly-model evaluation.
  • Alpha Net — Software Engineer Intern: JavaScript/AWS workflows and mobile REST features.

I use AI assistance during development and document project scope and limitations. Performance figures in project READMEs describe specified local workloads; they are not production traffic or user impact.

Pinned Loading

  1. chronicle-lsm-store chronicle-lsm-store Public

    Embedded C++20 LSM-tree engine with atomic WAL batches, skip lists, Bloom-filtered SSTables, recovery, range scans, and compaction.

    C++

  2. converge-crdt-board converge-crdt-board Public

    Local-first TypeScript collaboration board with CRDT conflict resolution, PostgreSQL persistence, and WebSocket sync.

    TypeScript

  3. queuepilot queuepilot Public

    Local support-routing rehearsal: capacity-limited review queues, reproducible NLP comparisons, versioned SQLite corrections, and verifiable evidence. No API keys.

    Python

  4. planclinic planclinic Public

    Interactive PostgreSQL JSON query-plan explorer with evidence-based performance clues, a visual execution tree, and browser-only analysis.

    JavaScript