🐦 Twitter Post Details

Viewing enriched Twitter post

@omarsar0

NEW research from IBM: Workflow Optimization for LLM Agents. LLM agent workflows involve interleaving model calls, retrieval, tool use, code execution, memory updates, and verification. How you wire these together matters more than most teams realize. This new survey maps the full landscape. It categorizes approaches along three dimensions: when structure is determined (static templates vs. dynamic runtime graphs), which components get optimized, and what signals guide the optimization (task metrics, verifier feedback, preferences, or trace-derived insights). It proposes structure-aware evaluation incorporating graph properties, execution cost, robustness, and structural variation. Most teams either hardcode their agent workflows or let them be fully dynamic with no principled middle ground. This survey provides a unified vocabulary and framework for deciding where your system should sit on the static-to-dynamic spectrum. Paper: https://t.co/qF8kTaNPYo Learn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX

Media 1
Media 2
View on Twitter

📊 Media Metadata

{
  "media": [
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2037536637954212332/media_0.png",
      "filename": "media_0.png"
    },
    {
      "type": "photo",
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2037536637954212332/media_1.png",
      "filename": "media_1.png"
    }
  ],
  "processed_at": "2026-03-27T14:36:08.938258",
  "pipeline_version": "2.0"
}

🔧 Raw API Response

{
  "type": "tweet",
  "id": "2037536637954212332",
  "url": "https://x.com/omarsar0/status/2037536637954212332",
  "twitterUrl": "https://twitter.com/omarsar0/status/2037536637954212332",
  "text": "NEW research from IBM: Workflow Optimization for LLM Agents.\n\nLLM agent workflows involve interleaving model calls, retrieval, tool use, code execution, memory updates, and verification.\n\nHow you wire these together matters more than most teams realize.\n\nThis new survey maps the full landscape.\n\nIt categorizes approaches along three dimensions: when structure is determined (static templates vs. dynamic runtime graphs), which components get optimized, and what signals guide the optimization (task metrics, verifier feedback, preferences, or trace-derived insights).\n\nIt proposes structure-aware evaluation incorporating graph properties, execution cost, robustness, and structural variation.\n\nMost teams either hardcode their agent workflows or let them be fully dynamic with no principled middle ground. This survey provides a unified vocabulary and framework for deciding where your system should sit on the static-to-dynamic spectrum.\n\nPaper: https://t.co/qF8kTaNPYo\n\nLearn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX",
  "source": "Twitter for iPhone",
  "retweetCount": 0,
  "replyCount": 1,
  "likeCount": 8,
  "quoteCount": 1,
  "viewCount": 609,
  "createdAt": "Fri Mar 27 14:26:04 +0000 2026",
  "lang": "en",
  "bookmarkCount": 9,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2037536637954212332",
  "displayTextRange": [
    0,
    279
  ],
  "inReplyToUserId": null,
  "inReplyToUsername": null,
  "author": {
    "type": "user",
    "userName": "omarsar0",
    "url": "https://x.com/omarsar0",
    "twitterUrl": "https://twitter.com/omarsar0",
    "id": "3448284313",
    "name": "elvis",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": null,
    "profilePicture": "https://pbs.twimg.com/profile_images/939313677647282181/vZjFWtAn_normal.jpg",
    "coverPicture": "https://pbs.twimg.com/profile_banners/3448284313/1565974901",
    "description": "",
    "location": "DAIR.AI Academy",
    "followers": 295553,
    "following": 795,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Fri Sep 04 12:59:26 +0000 2015",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 35214,
    "hasCustomTimelines": true,
    "isTranslator": true,
    "mediaCount": 4567,
    "statusesCount": 17531,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "2036819915097940174"
    ],
    "profile_bio": {
      "description": "Building @dair_ai • Prev: Meta AI, Elastic, PhD • New AI learning portal: https://t.co/1e8RZKs4uX",
      "entities": {
        "description": {
          "hashtags": [],
          "symbols": [],
          "urls": [
            {
              "display_url": "academy.dair.ai",
              "expanded_url": "https://academy.dair.ai/",
              "indices": [
                74,
                97
              ],
              "url": "https://t.co/1e8RZKs4uX"
            }
          ],
          "user_mentions": [
            {
              "id_str": "0",
              "indices": [
                9,
                17
              ],
              "name": "",
              "screen_name": "dair_ai"
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "dair.ai",
              "expanded_url": "https://www.dair.ai/",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/XQto5ypSIk"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.twitter.com/A9eMw7R8N5",
        "expanded_url": "https://twitter.com/omarsar0/status/2037536637954212332/photo/1",
        "ext_media_availability": {
          "status": "Available"
        },
        "features": {
          "large": {
            "faces": []
          },
          "orig": {
            "faces": []
          }
        },
        "id_str": "2037536634909208576",
        "indices": [
          280,
          303
        ],
        "media_key": "3_2037536634909208576",
        "media_results": {
          "id": "QXBpTWVkaWFSZXN1bHRzOgwAAQoAARxGyMYsm7AACgACHEbIxuIawewAAA==",
          "result": {
            "__typename": "ApiMedia",
            "id": "QXBpTWVkaWE6DAABCgABHEbIxiybsAAKAAIcRsjG4hrB7AAA",
            "media_key": "3_2037536634909208576"
          }
        },
        "media_url_https": "https://pbs.twimg.com/media/HEbIxiybsAAmmFd.png",
        "original_info": {
          "focus_rects": [
            {
              "h": 869,
              "w": 1552,
              "x": 0,
              "y": 0
            },
            {
              "h": 1552,
              "w": 1552,
              "x": 0,
              "y": 0
            },
            {
              "h": 1656,
              "w": 1453,
              "x": 0,
              "y": 0
            },
            {
              "h": 1656,
              "w": 828,
              "x": 0,
              "y": 0
            },
            {
              "h": 1656,
              "w": 1552,
              "x": 0,
              "y": 0
            }
          ],
          "height": 1656,
          "width": 1552
        },
        "sizes": {
          "large": {
            "h": 1656,
            "w": 1552
          }
        },
        "type": "photo",
        "url": "https://t.co/A9eMw7R8N5"
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "urls": [
      {
        "display_url": "arxiv.org/abs/2603.22386",
        "expanded_url": "https://arxiv.org/abs/2603.22386",
        "indices": [
          950,
          973
        ],
        "url": "https://t.co/qF8kTaNPYo"
      },
      {
        "display_url": "academy.dair.ai",
        "expanded_url": "https://academy.dair.ai/",
        "indices": [
          1026,
          1049
        ],
        "url": "https://t.co/1e8RZKs4uX"
      }
    ],
    "user_mentions": []
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "article": null
}