Mohit Goyal (Harness arc)

I built an AI that reads your research papers and never lies about what's in them. Meet MemexLLM ↓ • Upload 50+ formats (PDFs, YouTube, audio) • Chat → get cited answers with page numbers • Hybrid search finds ideas, not just keywords • Turns papers into podcasts 🎧 • Auto-generates flashcards & quizzes • 100% open source, self-hostable Built solo in 3 months. 📖 Case study: mohitx.in/projects/memex… ⭐ GitHub: github.com/MohitGoyal09/M… Demo

Got invitation from a journal to publish my research

Here is my past 2 weeks tokens usage on Hermes. i have tried various workflows/ special agent profiles to link with kaban board etc etc. Still not using it at full potential but it is fun!!

@Misbahtwts github.com/sponsors/Mohit…

@ByteMohit Share link brotha

Somebody lanuched something . Do support!! ❤️

@a_erdem4 Would like to join you. my profile - mohitx.in

Günün anlam ve önemine uygun bir şekilde bu sene mezun olacak ya da son 3 yıl içinde mezun olup AI üzerine çalışmak isteyen 8 kişiye Slack üzerinden mentörlük yapmayı düşünüyorum. Umarım LLMlere ya da reddite sorular sormaktan daha yararlı olur.

Winning together with your friends is next level Rizz!😊🤩 @__Frostbytee__ @ByteMohit @AnishaGarg85381

cookinnggg

Subagents added.

This year is all about Agentic Harness . so WHY NOT BUILD ONE YOURSELF.

We're drowning in AI tools, but starving for focus. So me and my friend built CogniFlow in 48 hours for the @runable_hq Build Challenge -> Syncs Google Cal + Apple Health to auto reschedule tasks by your energy levels -> Scans messy desks into checklists -> Rewrites robotic AI drafts in your brand voice Expo mobile + Hono web. @itsumeshk #RunableChallenge #VibeCoding

@Shreyas_Pandeyy bhai pata chle toh mujhe bhi btana , i have some ideas for "Aaas"

Question for builders working around runtime agents, OpenClaw, workflow orchestration, AI infra, and autonomous systems: Which industry is most underserved right now but will heavily need these tools soon? Trying to map where the strongest real-world demand is emerging beyond the hype.

is 280 skills too much ?

@JasgunChandnani true

@ByteMohit just one accepted changes the game

One more Rejection . i donot know when i am gonna get good in this game . 😭😭 Rejection Streak : 22x

Some Quick Updates on my harness: What i'm building: My own Agentic harness like for example claude code , opencode , pi etc. Why: Best way to learn something is doing it by first principles this what i am doing. IT IS JUST THAT SIMPLE. NO CAP. so, what is done till now and what is left to come . Features shipped ( Till now) : 1. All Primitive tools for an agent including grep , glop , edit file , wrtie file , read , todos , memory , web search , web fetch . 2. Good Interactive TUI. 3. Session & Config Management. What 's left to come : 1. Subagents ( currently in progress) 2. Agent hooks 3. Skill registery 4. Swarm Mode

@jaya_6400 true.

@ByteMohit It's tough to land a good company, and it takes effort. Hope your efforts get paid off 🙌🏻

@jaya_6400 efforts are all their . like me building things from scratch to understand better. hardly anybody pushing that hard.

@ByteMohit This is different I saw to post the rejection streak, u can trace back and count your efforts u gave while applying 😅

@scifi_tessa true . ikr

@ByteMohit Soft skills used to mean tolerating meetings. Now it's recognizing when AI is wrong and explaining why to stakeholders. Harder than it sounds.

The most underrated asset a human can have in 2026: the ability to sit in a room with another human, make eye contact, and have a real conversation. As AI handles more of the transactional stuff, the humans who can do the relational stuff become disproportionately valuable. The soft skills people used to dismiss as fluffy are becoming the hard skills. The hard skills people spent decades acquiring are becoming the soft ones.

Here is what i am following for next 6 months to become a GenAI Engineer. Stage 1: Python + Async Architecture FastAPI, asyncio, typing, Pydantic v2, event-driven design, API integration patterns. Stage 2: Multimodal LLM Fundamentals Transformer architecture, SLMs vs LLMs, context window management, vision/audio inputs, token economics. Stage 3: Structured Outputs + Prompt Ops JSON schema enforcement, function calling, prompt versioning, template management, few-shot optimization. Stage 4: Advanced RAG + Knowledge Graphs Hybrid search, graph RAG, semantic reranking, metadata filtering, citation grounding, incremental indexing. Stage 5: Agentic Workflows + Orchestration LangGraph/LlamaIndex, tool use, planning loops, multi-agent collaboration, human-in-the-loop handoffs. Stage 6: Production GenAI Applications Streaming responses, optimistic UI, fallback models, rate limiting, cost-aware routing, session management. Stage 7: Evaluation + Quality Assurance LLM-as-a-judge, automated eval harnesses, regression testing, hallucination metrics, RAGAS/DeepEval. Stage 8: Inference + Infrastructure Optimization vLLM/SGLang, quantization (FP8/INT4), KV caching, speculative decoding, edge deployment, model distillation. Stage 9: MLOps + Observability Distributed tracing, latency monitoring, cost dashboards, drift detection, CI/CD for prompts and models. Stage 10: AI Safety + Compliance Guardrails, prompt injection defense, PII redaction, copyright checks, EU AI Act compliance, content filtering. Stage 11: Open Source + Portfolio Ship multimodal agents publicly, write technical deep dives, record demo videos, contribute to orchestration libs. Stage 12: Apply GenAI Engineer, AI Application Developer, LLM Infrastructure Engineer, Autonomous Systems Engineer roles. Most people stay stuck watching tutorials. Builders get hired. (Bookmark it)

Running multiple audits on myself using hermes lately..

@Sankalp1906 Thanks. Let me fix it. I need to update my portfolio also. V2.1.01 releasing soon!!

@ByteMohit in ur portfolio resume isnt accessible