Steven

Building agents, workflows & automation

[Steven @recursiondealer]

@igmarinelli based in NYC open to roles! Lets connect

[AI, Ads & Apps  @AIAdsApps]

Here are 3 AI wrapper apps making $50K+/mo.

1. HairApp

[diva @divaagurlxw]

As an AI Engineer. Please learn

>Harness engineering, not just prompt engineering

>Context engineering, not just long prompts

>Prompt caching vs. semantic caching tradeoffs

>KV cache management, eviction, reuse, and memory pressure at scale

>Prefill vs. decode latency and why they optimize differently

>Continuous batching, paged attention, and throughput optimization

>Speculative decoding vs. quantization vs. distillation tradeoffs

>INT8, INT4, FP8, AWQ, GPTQ, and when quantization hurts quality

>Structured output failures, schema validation, repair loops, and fallback chains

>Function calling reliability, tool contracts, argument validation, and idempotency

>Agent guardrails, loop budgets, tool budgets, and termination conditions

>Model routing, graceful fallback logic, and degraded-mode UX

>RAG architecture: chunking, embeddings, hybrid search, reranking, and freshness

>Retrieval evals: recall, precision, grounding, attribution, and citation quality

>Evals: golden sets, regression tests, adversarial tests, LLM-as-judge, and human evals

>LLM observability as a first-class discipline: traces, spans, tokens, latency, errors, and drift

>Cost attribution per feature, workflow, tenant, and user journey not just per model

>Safety engineering: prompt injection defense, data leakage prevention, and permission boundaries

>Multi-tenant isolation, cache safety, and cross-user context contamination prevention

>Fine-tuning vs. in-context learning vs. RAG vs. distillation and when each is the wrong tool

>Latency, quality, cost, and reliability tradeoffs across the full inference stack

>Production failure modes: hallucinated tool calls, malformed JSON, stale retrieval, runaway agents, and silent eval regressions

[Meng To @MengTo]

I open-sourced the Kage landing page and the three.js skills behind it.

The whole thing is under 1mb of three.js and code, 3mb with the hq png images. For a site where every section scrolls through a 3D world, that's surprisingly good.

Live site: https://t.co/0c4euVBe6x

Repo: https://t.co/GqXcUiLmIS

Skills: https://t.co/9eEDISAqZx
- build-threejs-scroll-worlds
- falling-leaves
- pointer-trail-emitter

Btw this took a lot of tokens and prompts. A follow helps if you want more of these.

[Steven @recursiondealer]

@hiiinternet where at- DM me

[Steven @recursiondealer]

I’m looking to purchase a business asap. X connect me with brokers and lads that have fantastic deals. I’ve got 800k let’s talk

[Poonam @Aiwithpoonam]

A millionaire once told me about the "Cemetery Theory."

That was when I stopped procrastinating:

[Ridark @ridark_eth]

https://t.co/Zaf49OtXhb

[Priyanka Vergadia @pvergadia]

Research papers you must read for AI Engineer interviews:

1. Attention is all you need (Transformers)
2. LoRA (Low rank adaption)
3. PEFT ( Parameter Efficient Fine Tuning)
4. VIT (Vision Transformers) 5. VAE (Variational Auto Encoder)
6. GANs ( Generative Adversarial Networks)
7. BERT ( Bidirectional Encoder Representation from Transformers)
8. Diffusion Models (Stable Diffusion)
9. RAG (Retrieval Augment Generation)
10. GPT (Generative Pre-trained Transformers)

[Suraj Sharma @suraj_sharma14]

If I had 6 months to become a GenAI Engineer.

I'd do this.

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)

[Akshay 🚀 @akshay_pachaar]

A tricky LLM interview question:

Your RAG system scores 90% retrieval accuracy on 5k company docs.

But scaling to 500k docs drops the accuracy to just 50%, with the same embedding model and retriever.

Why did this happen?

The simplest answer is that more documents mean more competition for the top-k retrieval slots. That is true, but it doesn't explain why accuracy drops this dramatically.

The answer comes down to how enterprise docs are distributed in the embedding space.

Today, a single product decision in a company generates meeting transcripts, Slack threads, Confluence docs, Jira tickets, and email threads.

They are related to the same event, so they all land in a similar region of the embedding space.

As the company operates over months, this pattern repeats for every project/customer/roadmap, and the embedding space fills up with clusters of closely related documents.

But all related docs don't contain the same facts.

→ Slack thread covers the decision made
→ Jira has the implementation deadline
→ Confluence has the technical spec
→ Email thread has the customer request

When a query is about a specific fact (like a deadline), the answer lives in one of those docs.

At a 5K corpus size, there might be 3-5 docs touching that topic, and the correct one easily lands in the top-k results.

But at a 500K corpus size, there could be 40-60 total docs, and the one containing the actual answer can easily get pushed out of the top-k by other topically relevant docs, degrading retrieval.

A recent research paper from Onyx documented this.

The researchers used their newly open-sourced EnterpriseRAG-Bench dataset.

It has 500k+ synthetic enterprise documents spread across Slack, Gmail, Jira, GitHub, Confluence, Google Drive, HubSpot, Fireflies, and Linear, with realistic noise like misfiled documents, near-duplicates, and conflicting versions.

They ran the same retrievers at five corpus sizes from 5K to 500K.

→ Vector search accuracy dropped from 90.7% at 5K documents to 50.6% at 500K docs.

→ BM25 degraded more gracefully, from 85.8% to 68.4%.

→ At every scale, higher neighborhood density in the embedding space monotonically correlated with lower recall.

The practical implication here is that retrieval accuracy on a 5k test set tells you almost nothing about production-scale performance.

Always test at a realistic volume to measure the neighborhood density in your embedding space to estimate how much headroom the retriever actually has.

The entire EnterpriseRAG-Bench dataset (500K docs with questions, and the whole evaluation harness) is open-source.

Run your retriever against it at 5K, then at 500K, and see where your own accuracy curve breaks.

I have shared the GitHub repo in the replies.

[Ryan @WritesToProfit]

If I were in my 30s or 40s & wanted to retire in the next 10 years using AI, here’s exactly what I’d do:

1. Set up an LLC immediately. Not next month. Not after you "feel ready." This week.

[Steven @recursiondealer]

@c_valenzuelab Lets connect!

[Akshay Shinde @ConsciousRide]

90% of AI Agents interviews in 2026 are just these 10 concepts repeated:

[Steven @recursiondealer]

@boardyai Generative AI Engineer based in NYC- build multiple products production level and have a product experience! Lets connect DMs are open

[Akshay Shinde @ConsciousRide]

90% of AI Engineer interviews in 2026 are just these 12 concepts repeated:

[Steven @recursiondealer]

@hiiinternet send the invite

[Steven @recursiondealer]

@hiiinternet I use anthropic daily and use it to build production grade applications where I also red team them. Lets connect

[Steven @recursiondealer]

@hiiinternet looking for a role in generative AI if anyone has an opportunity. I've built a bunch I can showcase

[Dave Morin 🦞 @davemorin]

SOUL.md - who I am
MEMORY.md - what I've lived
DREAMS.md - what I'm becoming