Cities (264) ▾

AI Tinkerers - Post-Training

Filter
The image displays a presentation slide with a dark blue background, featuring a bold title about LLMs and synthetic data, accompanied by two interconnected diagrams illustrating data flow and business entity relationships. The slide includes speaker and company attribution at the bottom.. Text: ERA BY EON • RESEARCH LLMs Suck at Generating Synthetic Data one document at a time Slack 2027-03-14 CRM Closed Lost Email "signed" Ticket J. Park ? Jira no owner the company first Employee Deal Account Contact Call Ticket Salesforce Gong Zendesk Slack Benji Gruenbaum • Eon Era by Eon

LLMs Suck at Generating Synthetic Data

By Benji Gruenbaum · October 06, 2026 · 10 minutes · 4 photos

LLMs generate each synthetic document that reads fine, yet whole companies fall apart: in EnterpriseRAG, 60% of Slack messages were dated in the future and docs contradicted. Era fixes this by generating the entire company first from a resolved entity graph, then slicing consistent REST/MCP system views (60+ systems). Benchmarking nine models on one rerunnable company showed mean accuracy dropping from 92.6% (simple filtering) to 3.7% (multi-hop), with exact grading.

Read post →
The image displays a stylized world map with glowing network lines connecting various points, overlaid with numerous small rectangular boxes containing images of people and text labels. Prominently in the center, large blue text reads "AITINKERERS" and "POST-TRAINING", suggesting an informational or promotional graphic.. Text: AITINKERERS "POST-TRAINING" PALLADIUM ELENIOM BANDICAM GARDENWARE MOI AFARLE ENIGMA NO FARM IN RAIN CINE PLENIUR Mianny Hotel FREELANCE ALPHONSO S. Africa THE FUGA AL PRUUCE MAGNA COREM ABDUILE BERLLINE ME CEULANTEPIC AN [numerous other small, unreadable labels and names]

Top AI Demos #47: AI Video Editing, Agent Delete Gates, and Model Ethics

October 05, 2026 · 1 minute · 33 photos

This edition spotlights practical AI progress: AI video editing demos, “agent delete gates,” and model ethics discussions. If you build or test agents and media workflows, these are timely prompts to examine capabilities, control mechanisms, and responsible use—thanks for staying curious and sharing the community’s questions.

Read post →
The image displays a stylized world map with glowing network lines connecting various points, overlaid with numerous small rectangular boxes containing images of people and text labels. Prominently in the center, large blue text reads "AITINKERERS" and "POST-TRAINING", suggesting an informational or promotional graphic.. Text: AITINKERERS "POST-TRAINING" PALLADIUM ELENIOM BANDICAM GARDENWARE MOI AFARLE ENIGMA NO FARM IN RAIN CINE PLENIUR Mianny Hotel FREELANCE ALPHONSO S. Africa THE FUGA AL PRUUCE MAGNA COREM ABDUILE BERLLINE ME CEULANTEPIC AN [numerous other small, unreadable labels and names]

Top AI Demos #45: Live Agent Intent Streaming Memory and Agent Sandboxing

September 21, 2026 · 1 minute · 36 photos

Top AI Demos #45 shares live agent intent streaming built for memory-aware agents, plus a practical sandbox approach for safer agent execution. If you’re building or testing conversational agents, these demos show how to stream intents in real time while keeping state and running code in an isolated environment.

Read post →
An abstract digital illustration featuring a network of interconnected purple lines and circles on a very light lavender background. A prominent central node contains a stylized white 'R' symbol within a dark purple circle, suggesting a logo or a key element in the network.

How to Build Antifragile Agents with OpenRouter

By Kenny Rogers · August 05, 2026 · 16 minutes · 2 photos

After Anthropic suspended Fable 5 (June 12–July 1, 2026), we built an antifragile billing agent that keeps working when models/providers fail by routing through OpenRouter presets and evaluated fallbacks. Using TypeScript code for money/policy logic and a 20-case synthetic eval (8 holdout), Grok 4.5 matched all decomposed cases (20/20) at 63% lower cost than Fable 5 ($0.0555 vs $0.1492) and Grok became primary while Fable stayed fallback. We store model order in an OpenRouter preset (no deploy needed) and use Router metadata/Broadcast to trace which model answered.

Read post →
A close-up photograph shows a person's hands typing on a laptop, which displays a website for 'Anam' featuring interactive avatars for sales coaching. In the blurred background, several other individuals are visible working at tables in what appears to be an office or co-working space.. Text: Anam Anam - AI Avatars for Sales anam.ai Anam Product Solutions Resources Pricing Sign in Sign up Meet our most expressive model: Cara Create interactive avatars for 'sales coaching'. Sign up Get a demo Anam SFOS SF Mac M Yelp

What I Learned Giving Fable 5 a Face

By Ben Carr · July 30, 2026 · 11 minutes · 4 photos

Giving an avatar “a face” bottlenecks on interaction timing, not rendering: interruptions, turn-taking/ending cues (e.g., Turkish needed eager end-of-turn disabled), visible thinking time, and true audio-driven lip sync. With Anam bring-your-own-LLM, keep the system prompt short (1–2 spoken sentences) and stream Fable 5 tokens into the talk stream; Anam targets ~180ms response.

Read post →
An infographic diagram illustrating the 'REACT LOOP' framework, divided into three categories: Skills, Agent types, and Runtime constraints.. Text: Primitives of Reef skills - agent types - runtime constraints REACT LOOP Skills • Procedural instructions • Principles over N-shot examples • Python bindings w/ decorators — not separate schemas • In-process invocations — not bash subprocess Agent types • Parallel & sequential execution • Planner for routing • Specialists for compute Runtime constraints • Anti-hallucination constraints • Convergence constraints • Data access constraints Architect your own domain specific framework

How to Write a Winning Agent Harness for Your Domain

By Hitesh Jain · June 24, 2026 · 13 minutes

If adding instructions makes your agent worse, it’s often a harness problem: we watched Reef/AlphaCumen collapse from a Vals v1 lead to zero on Vals v2 because prompts exceeded 2,500 lines and 50+ overlapping tools, causing token-budget timeouts. Reef fixes this with git-native, skills-first ReAct: skills bound by `SKILL.md` + `@skill_fn`, lazy-loaded skill indexes, planner specialist routing, and runtime constraints like `inject|clamp` as-of cutoffs. Result on finance evals (n=239): 82.6% pass rate with full structure vs 44.87% reference and 49.8% retrieval-only, at ~10× lower cost (620k tokens/query; $0.13 vs $1.35 at Opus 4.7).

Read post →
A laptop sits open on a light wooden desk next to a mug, displaying the text 'AI TINKERERS LOCAL LLMS ON MAC' on its screen.. Text: AI TINKERERS LOCAL LLMS ON MAC

How to Run Open-Source LLMs Locally on a Mac with MLX-LM

By Eric Fillion · June 12, 2026 · 8 minutes

Learn to run open-source LLMs locally on Mac Apple Silicon with Apple’s MLX-LM: install `pip install mlx-lm`, then `load()` a Hugging Face model and call `generate()` (optionally `stream_generate()`); examples show ~35.5 tok/s on an Air (smollm3-3b-mlx, ~2.6 GB peak) and ~79 tok/s on M4 Max (gpt-oss-120b-mlx, ~63 GB peak).

Read post →
The image displays a stylized world map with glowing network lines connecting various points, overlaid with numerous small rectangular boxes containing images of people and text labels. Prominently in the center, large blue text reads "AITINKERERS" and "POST-TRAINING", suggesting an informational or promotional graphic.. Text: AITINKERERS "POST-TRAINING" PALLADIUM ELENIOM BANDICAM GARDENWARE MOI AFARLE ENIGMA NO FARM IN RAIN CINE PLENIUR Mianny Hotel FREELANCE ALPHONSO S. Africa THE FUGA AL PRUUCE MAGNA COREM ABDUILE BERLLINE ME CEULANTEPIC AN [numerous other small, unreadable labels and names]

Top AI Demos #30: Clinical Co-Pilots, Pull Agents, and Life Planners

June 08, 2026 · 1 minute · 81 photos

This roundup highlights three AI directions: clinical co-pilots for healthcare workflows, “pull agents” that proactively retrieve needed information, and life planners that translate goals into actionable plans. Thanks to everyone building these tools—share your real-world wins and what’s working.

Read post →
The image displays a stylized world map with glowing network lines connecting various points, overlaid with numerous small rectangular boxes containing images of people and text labels. Prominently in the center, large blue text reads "AITINKERERS" and "POST-TRAINING", suggesting an informational or promotional graphic.. Text: AITINKERERS "POST-TRAINING" PALLADIUM ELENIOM BANDICAM GARDENWARE MOI AFARLE ENIGMA NO FARM IN RAIN CINE PLENIUR Mianny Hotel FREELANCE ALPHONSO S. Africa THE FUGA AL PRUUCE MAGNA COREM ABDUILE BERLLINE ME CEULANTEPIC AN [numerous other small, unreadable labels and names]

Top AI Demos #29: Agent Memory, Image Relighting & Software Testing

June 01, 2026 · 1 minute · 73 photos

In this week’s top AI demos (#29), explore three practical highlights: agent memory for smarter ongoing context, image relighting for dynamic lighting changes, and software testing workflows that help validate results faster. Thanks to everyone sharing these demos—great ideas to try and build on.

Read post →
A promotional banner for 'homecrew', an open-source package manager for agent skills, featuring its logo and a terminal interface mock-up.. Text: homecrew An open-source package manager for agent skills ~/work · zsh $ crew install founding-engineer ✓ synced across Claude Code, Codex, Gemini $ crew update ✓ team skills are current

Homecrew: An Open-Source Package Manager For Agent Skills

By Steve Krenzel · May 12, 2026 · 7 minutes

Homecrew is an open-source package manager for agent skills: a skill is a directory with a `SKILL.md`, and a tap is a git repo or local directory of skills. Install with `crew tap add …` + `crew install`, then `crew update` keeps skills synced across all detected agents (e.g., `~/.agents/skills/` and agent-specific dirs), including dependency installs and automatic tap-level updates via `crew autoupdate enable`.

Read post →
The image displays a stylized world map with glowing network lines connecting various points, overlaid with numerous small rectangular boxes containing images of people and text labels. Prominently in the center, large blue text reads "AITINKERERS" and "POST-TRAINING", suggesting an informational or promotional graphic.. Text: AITINKERERS "POST-TRAINING" PALLADIUM ELENIOM BANDICAM GARDENWARE MOI AFARLE ENIGMA NO FARM IN RAIN CINE PLENIUR Mianny Hotel FREELANCE ALPHONSO S. Africa THE FUGA AL PRUUCE MAGNA COREM ABDUILE BERLLINE ME CEULANTEPIC AN [numerous other small, unreadable labels and names]

⚡ See 24/7 Dev Agents + MCP Apps

April 20, 2026 · 1 minute · 59 photos

⚡ Explore 24/7 Dev Agents paired with MCP Apps—always-on agents that keep working and MCP apps that integrate smoothly—so your workflow stays active around the clock. Thanks for building with us and sharing what you’re testing!

Read post →
The image displays a stylized world map with glowing network lines connecting various points, overlaid with numerous small rectangular boxes containing images of people and text labels. Prominently in the center, large blue text reads "AITINKERERS" and "POST-TRAINING", suggesting an informational or promotional graphic.. Text: AITINKERERS "POST-TRAINING" PALLADIUM ELENIOM BANDICAM GARDENWARE MOI AFARLE ENIGMA NO FARM IN RAIN CINE PLENIUR Mianny Hotel FREELANCE ALPHONSO S. Africa THE FUGA AL PRUUCE MAGNA COREM ABDUILE BERLLINE ME CEULANTEPIC AN [numerous other small, unreadable labels and names]

⚡ Autonomous Software Factories + Agentic Security

April 13, 2026 · 1 minute · 25 photos

⚡ Autonomous software factories pair well with agentic security to help teams ship faster with safer builds. The key idea: use agents to automate parts of the software pipeline while applying security checks along the way, so reliability and protection stay built-in as work progresses.

Read post →
The image displays a stylized world map with glowing network lines connecting various points, overlaid with numerous small rectangular boxes containing images of people and text labels. Prominently in the center, large blue text reads "AITINKERERS" and "POST-TRAINING", suggesting an informational or promotional graphic.. Text: AITINKERERS "POST-TRAINING" PALLADIUM ELENIOM BANDICAM GARDENWARE MOI AFARLE ENIGMA NO FARM IN RAIN CINE PLENIUR Mianny Hotel FREELANCE ALPHONSO S. Africa THE FUGA AL PRUUCE MAGNA COREM ABDUILE BERLLINE ME CEULANTEPIC AN [numerous other small, unreadable labels and names]

⚡ Agent Version Control & Pipeline-Parallel LLMs

March 30, 2026 · 1 minute · 68 photos

Agent version control plus pipeline-parallel LLMs help teams iterate safely while scaling throughput. By tracking changes across agent versions and running models in parallel stages, you reduce risk and speed up experimentation—so workflows stay consistent as you push new ideas into production.

Read post →
A small, white, rounded robot with two black circular eyes and thin wire antennas, sitting on a desk next to a stack of books.. Text: SWITZERLAND SCHWEIZ-SUISSE

What I Learned Deploying OpenClaw Beyond Demos

By Kaya Jones · March 15, 2026 · 11 minutes · 2 photos

After three months deploying OpenClaw beyond demos, the workflows that stick follow one shape: live inside tools you already use, carry context forward, and save time without babysitting. I built four systems—Robin’s nightly arXiv research (silence on weak results), a two-layer knowledge system (Obsidian facts + Honcho decision reasoning queried together), SnapshotClaw versioning for ~/.openclaw (configs/cron/skills/memory/plugins) to recover after 2AM breakage, and a Reachy Mini “brain” via Botwick that made the robot useful the next day (mic→VAD→STT→Claude→TTS).

Read post →
The image displays a stylized world map with glowing network lines connecting various points, overlaid with numerous small rectangular boxes containing images of people and text labels. Prominently in the center, large blue text reads "AITINKERERS" and "POST-TRAINING", suggesting an informational or promotional graphic.. Text: AITINKERERS "POST-TRAINING" PALLADIUM ELENIOM BANDICAM GARDENWARE MOI AFARLE ENIGMA NO FARM IN RAIN CINE PLENIUR Mianny Hotel FREELANCE ALPHONSO S. Africa THE FUGA AL PRUUCE MAGNA COREM ABDUILE BERLLINE ME CEULANTEPIC AN [numerous other small, unreadable labels and names]

AI Tinkerers #16: Vector DB Evolution & Persistent Agent Memory

February 23, 2026 · 1 minute · 48 photos

AI Tinkerers #16 dives into vector database evolution and persistent agent memory. The session highlights practical shifts in how embeddings are stored and retrieved, plus strategies for keeping agent context across sessions—so builders can design more reliable, stateful AI systems.

Read post →

Post-Training Fellows

Apply to become a Fellow

Have a bold AI idea, credible proof, and a hard-won lesson serious builders can use? Pitch a practical Deep Dive for 130K+ leading AI practitioners globally.

Apply to become a Fellow
Events Gallery FAQ Sponsor