🐦 Twitter Post Details

Viewing enriched Twitter post

@sh_reya

One of the most pressing questions in our AI Evals course is: "Why can’t I just have an LLM write my LLM pipeline?" The nuanced answer is that you can use LLMs to assist, but not for the whole pipeline. Knowing where to put the LLM in the loop is the hard part. To unpack this, we invited Omar Khattab (@lateinteraction) —creator of DSPy, leading expert on prompt optimization, and now professor at MIT—for a "fireside chat" in the course. He shed light on how he approaches pipeline development in practice. What stood out to us is that Omar spends most of his time on specification—e.g., defining the task clearly, looking at the data, and doing careful error analysis—before letting LLMs automate anything. This up-front rigor is what makes downstream optimization actually work. We've put the recording on YouTube. If you're wondering how Omar thinks about these tradeoffs, this conversation is worth a listen! https://t.co/j3D83hRLKW

Media 1

📊 Media Metadata

{
  "media": [
    {
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1961110090314125524/media_0.jpg?",
      "media_url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/1961110090314125524/media_0.jpg?",
      "type": "photo",
      "filename": "media_0.jpg"
    }
  ],
  "processed_at": "2025-08-29T00:14:27.025998",
  "pipeline_version": "2.0"
}

🔧 Raw API Response

{
  "type": "tweet",
  "id": "1961110090314125524",
  "url": "https://x.com/sh_reya/status/1961110090314125524",
  "twitterUrl": "https://twitter.com/sh_reya/status/1961110090314125524",
  "text": "One of the most pressing questions in our AI Evals course is: \"Why can’t I just have an LLM write my LLM pipeline?\" The nuanced answer is that you can use LLMs to assist, but not for the whole pipeline. Knowing where to put the LLM in the loop is the hard part.\n\nTo unpack this, we invited Omar Khattab (@lateinteraction) —creator of DSPy, leading expert on prompt optimization, and now professor at MIT—for a \"fireside chat\" in the course. He shed light on how he approaches pipeline development in practice. What stood out to us is that Omar spends most of his time on specification—e.g., defining the task clearly, looking at the data, and doing careful error analysis—before letting LLMs automate anything. This up-front rigor is what makes downstream optimization actually work.\n\nWe've put the recording on YouTube. If you're wondering how Omar thinks about these tradeoffs, this conversation is worth a listen! https://t.co/j3D83hRLKW",
  "source": "Twitter for iPhone",
  "retweetCount": 18,
  "replyCount": 8,
  "likeCount": 112,
  "quoteCount": 5,
  "viewCount": 21239,
  "createdAt": "Thu Aug 28 16:54:15 +0000 2025",
  "lang": "en",
  "bookmarkCount": 122,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "1961110090314125524",
  "displayTextRange": [
    0,
    279
  ],
  "inReplyToUserId": null,
  "inReplyToUsername": null,
  "author": {
    "type": "user",
    "userName": "sh_reya",
    "url": "https://x.com/sh_reya",
    "twitterUrl": "https://twitter.com/sh_reya",
    "id": "2286218053",
    "name": "Shreya Shankar",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": null,
    "profilePicture": "https://pbs.twimg.com/profile_images/1342529111839944705/hzr44mb5_normal.jpg",
    "coverPicture": "",
    "description": "",
    "location": "Berkeley, CA",
    "followers": 47765,
    "following": 683,
    "status": "",
    "canDm": true,
    "canMediaTag": true,
    "createdAt": "Sat Jan 11 06:46:16 +0000 2014",
    "entities": {
      "description": {
        "urls": []
      },
      "url": {}
    },
    "fastFollowersCount": 0,
    "favouritesCount": 14503,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 389,
    "statusesCount": 5447,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "1838617833393283428"
    ],
    "profile_bio": {
      "description": "doing a PhD @Berkeley_EECS, building https://t.co/PmuOqAYt6q | teaching https://t.co/CTWJ6z0JEg | formerly ML eng & undergrad @Stanford CS",
      "entities": {
        "description": {
          "urls": [
            {
              "display_url": "docetl.org",
              "expanded_url": "http://docetl.org",
              "indices": [
                37,
                60
              ],
              "url": "https://t.co/PmuOqAYt6q"
            },
            {
              "display_url": "bit.ly/evals-ai",
              "expanded_url": "http://bit.ly/evals-ai",
              "indices": [
                72,
                95
              ],
              "url": "https://t.co/CTWJ6z0JEg"
            }
          ],
          "user_mentions": [
            {
              "id_str": "0",
              "indices": [
                12,
                26
              ],
              "name": "",
              "screen_name": "Berkeley_EECS"
            },
            {
              "id_str": "0",
              "indices": [
                126,
                135
              ],
              "name": "",
              "screen_name": "Stanford"
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "sh-reya.com",
              "expanded_url": "http://www.sh-reya.com",
              "indices": [
                0,
                23
              ],
              "url": "https://t.co/j0wChge195"
            }
          ]
        }
      }
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {},
  "card": null,
  "place": {},
  "entities": {
    "urls": [
      {
        "display_url": "youtube.com/watch?v=ctyU0z…",
        "expanded_url": "https://www.youtube.com/watch?v=ctyU0zfWgrA&themeRefresh=1",
        "indices": [
          917,
          940
        ],
        "url": "https://t.co/j3D83hRLKW"
      }
    ],
    "user_mentions": [
      {
        "id_str": "1605274291569799168",
        "indices": [
          304,
          320
        ],
        "name": "Omar Khattab",
        "screen_name": "lateinteraction"
      }
    ]
  },
  "quoted_tweet": null,
  "retweeted_tweet": null,
  "isLimitedReply": false,
  "article": null
}