Dharmendra Valiya

Senior Full-Stack, Mobile & AI Engineer | 9+ Years | React Native, iOS/Swift, TypeScript, React/Next.js, Node.js, Python | AI Agents | 2x Founding Engineer

[Dharmendra Valiya @dvvaliya]

hiring and need someone fast? skip the 5 interview rounds and run a short trial. fork a repo, give limited access, hand over one real task. you see how they ship, they see if they vibe with the product.

i've joined a couple of companies this way. open to it again, let's connect

[Dharmendra Valiya @dvvaliya]

react native tip from shipping apps past 1M users: test on the cheapest android phone your users actually have. the simulator on your laptop is lying to you about scroll jank

[Dharmendra Valiya @dvvaliya]

@lintaho Interested! I’ve spent 9+ years building, most recently a remote agent system that runs 12 agents in parallel.
I owned the secure sandbox execution and real-time streaming, from design to production.

Just sent you a DM

[Dharmendra Valiya @dvvaliya]

if your agent can't tell you what it actually changed, it's not done. tests passing isn't a receipt. the diff, the commands it ran and anything it touched outside the repo are

[Dharmendra Valiya @dvvaliya]

@eeshakhanna1 Interested! I’ve spent 9+ years building, most recently a remote agent system at Blackbox AI that runs 12 agents in parallel.
I owned the secure sandbox execution and real-time streaming, from design to production.

Just sent you a DM with more details

[Dharmendra Valiya @dvvaliya]

a new #1 model every two weeks means #1 stopped meaning much. the useful question is how fast your team can swap models without rewriting prompts and evals

[Dharmendra Valiya @dvvaliya]

proof of work: built Remote Agent at Blackbox AI (30M+ users), running untrusted AI-generated code for many users across up to 12 agents in isolated sandboxes, plus the Agent API. before that, a self-custody Bitcoin multisig wallet at Theya. I also post about agent safety + AI dev tooling here daily. would love to chat

[Dharmendra Valiya @dvvaliya]

not a PM, but if Moss needs an engineer to build the launch + trading products: built a self-custody Bitcoin multisig wallet at Theya, Stripe payments in production, and agent infra at Blackbox AI (30M+ users: Remote Agent, Agent API). founding engineer twice, available now. would love to help build it

[Dharmendra Valiya @dvvaliya]

watch out for agents that make a failing test pass by editing the assertion. it shows up as a green build and it's really a deleted check. worth diffing test files separately in every agent PR. anyone have a lint rule for this?

[Dharmendra Valiya @dvvaliya]

@CamiloVlies thanks Camilo, appreciate you checking with them. I'm open to relocating to SF with visa sponsorship, so on-site works if they'd sponsor. and if remote/contract opens up, I'm in. available immediately

[Dharmendra Valiya @dvvaliya]

infra side here. built Remote Agent at Blackbox AI (30M+ users): one task fans out to up to 12 agents in parallel sandboxes, results pushed to GitHub, plus the Agent API on top. running lots of isolated jobs in parallel and verifying their output is basically what a research factory needs. available now, my DMs are open (X is blocking my new DMs)

[Dharmendra Valiya @dvvaliya]

been job hunting for 2 months. what's worked: replying to founders' hiring posts with one specific thing I built. what hasn't: polished cover letters nobody reads. if you're hiring for agents, React Native or full stack, my DMs are open

[Dharmendra Valiya @dvvaliya]

@andotvu tried to DM but X is blocking new messages from my side. would love to teach this: built Remote Agent + the Agent API at Blackbox AI (30M+ users), founding engineer twice, and tech lead shipping a healthcare platform to 20K+ users. happy to travel for cohorts. my DMs are open

[Dharmendra Valiya @dvvaliya]

tried to DM but X is blocking new messages from my side. interested! 9+ yrs frontend-heavy: React/Next.js/TS, React Native apps past 1M users, and Blackbox Claw (Vercel-like AI dev platform, real-time streaming UI) at Blackbox AI. available immediately. my DMs are open, would love to share my CV

[Dharmendra Valiya @dvvaliya]

interested in the full stack role. DB to UI is my home turf: founding engineer twice, React/Next.js/Node/Postgres, React Native apps past 1M users, Stripe payments in production, and most recently agent infra at Blackbox AI (30M+ users). remote, available immediately. DMs open, would love to chat

[Dharmendra Valiya @dvvaliya]

built Remote Agent at Blackbox AI (30M+ users): one coding task fans out to up to 12 agents (Claude, Codex, Gemini, Grok) in parallel sandboxes, results pushed to GitHub.

hardest bug: agents confidently reporting "done" on code that didn't even run. the agent's own report was the bug. fix was to stop trusting self-reports and judge the actual output in the sandbox before anything touched the user's repo.

remote, available now, DMs open

[Dharmendra Valiya @dvvaliya]

just sent you mine, kept it to things you can verify:

- founding engineer twice, 9+ yrs end to end
- built Remote Agent at Blackbox AI (30M+ users): one task, up to 12 coding agents in parallel sandboxes, results on GitHub
- RN/iOS apps past 1M users, healthcare platform for 20K+

happy to do a paid build task if the founder wants proof over words

[Dharmendra Valiya @dvvaliya]

raising my hand for this one. technical co-founder who turns tools into products people actually use:

- founding engineer twice, 9+ yrs shipping 0 to 1
- built Remote Agent at Blackbox AI (30M+ users): one task fans out to up to 12 coding agents (Claude, Codex, Gemini...) in parallel sandboxes. so yes, i run a small army of Claude plans daily
- shipped React Native/iOS apps past 1M users, Stripe payments in production, Next.js/Node/Python/Postgres

raising capital: built alongside founders through early rounds, haven't led one myself yet
distribution: have shipped to 30M+ and 1M+ user bases, know what makes people stick
https://t.co/2alGmkhuMU

open to equity, available immediately, DMs open

[Dharmendra Valiya @dvvaliya]

@jpbassinello the confidence score is a smart touch. "remote" with 40% confidence vs 95% is exactly the signal i'd want before spending an hour on an application. will give it a proper try this week

[Dharmendra Valiya @dvvaliya]

@pawelk411 @maxjendrall lazy as in stopping early and calling it done? that one's usually fixable with a "verify before you finish" step in the prompt

[Dharmendra Valiya @dvvaliya]

@mehulmpt curious how much of that is context size. long sessions resend the whole history every turn, so a fresh thread per task usually stretches the limit more than switching models does