grobat79

Location: Kolkata, India (UTC+5:30)

Remote: Yes, remote-first, comfortable working async across timezones

Willing to relocate: Yes

Technologies: Rust, TypeScript, C#/.NET, distributed systems, network protocols (gRPC, SSZ-REST, binary transports), MCP, ZK (Groth16/Arkworks, RISC Zero)

Résumé/CV: https://github.com/Dyslex7c

Email: [email protected]

Final-year CS student (graduating mid-2027), available immediately for internship, contract, part-time or full-time roles. Looking for roles in AI agent infrastructure and runtimes, distributed backend, or protocol and consensus engineering.

Recent work:

- Nethermind (Ethereum client, C#), SWE intern, Mar-Jun 2026: worked on Pureth (trustless log and state indexing). Built an SSZ-REST binary transport that cut latency 4.8x vs JSON-RPC, and did a type-unification refactor (long/UInt256 to ulong) across ~1,300 files, reviewed by core engineers. Also fixed failing state-transition spec tests.

- Grandine (Rust consensus client): integrated the Ziren zkVM for off-chain execution verification (merged PR #385).

- LayerEdge, SDE intern: built a Rust service on the RISC Zero zkVM, with Merkle-tree state-change detection to avoid redundant ZK proof verification.

- Spigot (open source): an MCP control plane and per-call billing gateway for autonomous agents, with tool-call authorization, budget caps, atomic ledgering and idempotency. Live: https://www.spigot.sh/ | Repos: https://github.com/orgs/Spigot-SH/repositories

- cadas-core (Rust): BFT consensus research prototype with fault-injection simulation.

Also contributed to Reth, Besu, Ream and Forest (PR links in my GitHub profile README). Won the Unfold 2024 hackathon's Bullieverse AI Agent track.

I like turning specs into working code and finding spec-vs-implementation bugs.

Location: Kolkata, India (UTC+5:30)

Remote: Yes, remote-first (open to work in any timezone, comfortable working async with distributed teams)

Willing to relocate: Yes

Technologies: TypeScript, Rust, C#/.NET, distributed systems, network protocols (gRPC, binary transports, REST), Model Context Protocol (MCP), ZK (Groth16 on BLS12-381, Arkworks, RISC Zero zkVM)

Résumé/CV: https://github.com/Dyslex7c

Email: [email protected]

Final-year CS student (graduating mid-2027) working on distributed systems, network protocols, and AI agent infrastructure. Looking for remote roles in AI/ML systems, agent runtimes, or distributed backend/protocol engineering. Open to work across any timezone.

Recent work:

- Spigot: Building an HTTP per-call micro-billing gateway and Model Context Protocol (MCP) server for autonomous AI agents, handling tool call authorization, atomic ledgering, and idempotency.

- Nethermind (Ethereum runtime client, C#): did an SWE internship (Mar-Jun 2026), and worked on Pureth (trustless log and state indexing engine). Implemented witness capture endpoints and built an SSZ-REST binary transport that cut latency by 4.8x over JSON-RPC. Also led a type-unification refactor (`long`/`UInt256` to `ulong`) across ~1300 files reviewed by core engineers, and fixed failing state-transition specification tests.

- Grandine (Rust consensus client): integrated the Ziren zkVM into Grandine (merged PR #385) for off-chain execution verification.

- LayerEdge (SDE intern, Singapore): built a RISC Zero zkVM-backed Rust service and Merkle-tree state-change detection to cut redundant ZK proof verification.

- cadas-core (Rust research prototype): DispersedSimplex BFT consensus with FRIDA erasure coding, fault-injection simulation, and byte-level network metrics.

- ZK from scratch: Groth16 zk-SNARK on BLS12-381 (Arkworks); Chaum-Pedersen ZKP gRPC system in Rust with Miller-Rabin prime generation and Fiat-Shamir.

Also contributed to Reth, Besu, Ream, Forest, and distributed telemetry tooling. Won the Unfold 2024 hackathon Bullieverse AI Agent track.

Looking for opportunities in the AI/ML space (AI agent infrastructure, MCP tool execution, runtime systems), distributed systems, zkVMs, and consensus/protocol engineering. I like turning specs into working code and finding spec-vs-implementation bugs.

zk-Election is a groundbreaking decentralized voting framework that ensures privacy, security, and transparency using zk-SNARKs. Built with cutting-edge cryptographic techniques, it enables voters to cast their ballots anonymously while guaranteeing the integrity of election outcomes. With Circom-based circuits for efficient proof generation, on-chain verification via Solidity, and a Next.js-powered frontend, zk-Election eliminates the need for trust in election systems, providing a scalable, tamper-proof, and privacy-preserving solution for modern governance.

A smart agriculture web application aimed at providing farmers with data-driven insights using advanced machine learning, rainfall analysis, crop recommendations, and disease prediction

thanks man I appreciate it but that government thing was too far-fetched I think

It costed around a dollar. I don't know if that's gas efficient because I'm still a beginner in this domain, started exploring web3 just a month ago. What I did to make it gas efficient was to upload image/video files to Pinata as soon as the user attaches them and then later when the user posts them, the generated hash gets sent to the blockchain node. And talking about saved posts being paid for in Pinata I haven't really thought about it since I myself used the pinata free tier for this.

And noted down that nature of Ethereum, will verify it and update the details later. Thanks for pointing it out.