Grouping of records in case values are null

Each change in any parameter generates a new series of values ("dataSeries") for all parameters.
If several parameters are changed at the same time, we combine them into one series.

We take distinct values of "start_ts" for every "master_work_ordr_id".
This "dataSeries" is actual up to next change. Calculate this "up_to_ts" as next_ts-1 mcs like

  dateadd(mcs,-1,lead(start_ts,1,'9999-12-31 23:59:59.999999')
          over(partition by master_work_ordr_id order by start_ts)) up_to_ts

We are creating such a "framework" of the report.

 select master_work_ordr_id,start_ts 
        ,dateadd(mcs,-1,lead(start_ts,1,'9999-12-31 23:59:59.999999')
          over(partition by master_work_ordr_id order by start_ts)) up_to_ts
 from (select distinct master_work_ordr_id,start_ts from WORK_ORDR_IDENTIFIER_HIST)x

Then, for each "dataSeries", we get the current value of all the parameters. To do this, attach (LEFT JOIN) the "WORK_ORDR_IDENTIFIER_HIST" table to the framework.

For source data

Finally, PIVOT data

with dataSeries as(
  select master_work_ordr_id,start_ts 
        ,dateadd(mcs,-1,lead(start_ts,1,'9999-12-31 23:59:59.999999')
          over(partition by master_work_ordr_id order by start_ts)) up_to_ts
  from (
       select distinct master_work_ordr_id,start_ts 
       from WORK_ORDR_IDENTIFIER_HIST
       where MASTER_WORK_ORDR_ID = '955.244.397'
      )x
)
, expData as(
  select  ds.master_work_ordr_id,ds.start_ts,ds.up_to_ts,t.end_ts
    ,t.WORK_ORDR_ID_CTXT, t.WORK_ORDR_NUM
  from dataSeries ds
  left join WORK_ORDR_IDENTIFIER_HIST t on ds.master_work_ordr_id=t.master_work_ordr_id
  and ds.start_ts between t.start_ts and t.end_ts 
  and t.MASTER_WORK_ORDR_ID = '955.244.397'
)
SELECT MASTER_WORK_ORDR_ID WORK_ORDR_ID 
    ,START_TS,up_to_ts  -- ,"UTS" UTS ,"SER" SER ,"SPRINT" SPRINT ,"LDB" LDB 
    ,"NPA" NPA ,"WFM_INTERNAL" WFM_INTERNAL -- ,"INTERNAL" INTERNAL ,"ABDTBR" ABDTBR ,"WFMIL" WFMIL 
    ,"WFM" WFM 
    ,"CALLID" CALLID 
FROM 
    (
    SELECT  MASTER_WORK_ORDR_ID ,START_TS,up_to_ts ,WORK_ORDR_ID_CTXT ,WORK_ORDR_NUM
    FROM expData 
--    WHERE 1=1 AND MASTER_WORK_ORDR_ID = '955.244.397' 
    )   s_tab 
    PIVOT   
    (
        max(WORK_ORDR_NUM) 
        FOR WORK_ORDR_ID_CTXT IN
            ( -- "UTS" ,"SER" ,"SPRINT" ,"LDB"
                 "NPA" ,"WFM_INTERNAL"  --    ,"INTERNAL" ,"ABDTBR" ,"WFMIL"
                ,"WFM" ,"CALLID" 
            ) 
    ) p_tab 
ORDER BY MASTER_WORK_ORDR_ID, START_TS;     
;

Intermediate result of JOIN dataSeries and WORK_ORDR_IDENTIFIER_HIST see in fiddle.

select *
from(  select distinct master_work_ordr_id,start_ts 
        ,dateadd(mcs,-1,lead(start_ts,1,'9999-12-31 23:59:59.999999')
          over(partition by master_work_ordr_id order by start_ts)) up_to_ts
      from WORK_ORDR_IDENTIFIER_HIST) ds
left join WORK_ORDR_IDENTIFIER_HIST t on ds.master_work_ordr_id=t.master_work_ordr_id
  and ds.start_ts between t.start_ts and t.end_ts
;

Next time (up_to_ts) calculation example

select *
  ,cast(dateadd(mcs,-1,dttm) as datetime2(7)) res
from( select cast('2025-06-02 12:31:21.2004100' as datetime2(7)) dttm)t

For "master_work_ordr_id" data type you use "955.244.397" as string (varchar) and as integer (MASTER_WORK_ORDR_ID = 955244397). I gave the example as varchar. This does not affect the decision.

fiddle

I'm going to assume a few things...

1. There are actually no overlaps in the source data (there are none in your post)
2. Therefore there are no overlaps in the output data (so, your post must be mistaken)
3. There isn't actually a 1µs gap if a row ends at .407642 followed by a row at .407643

As such, my expected results for your sample data would be...

First...Things are significantly easier if time ranges are "inclusive start, exclusive end". That is 12:05:06.001 to 12:06:06.001 would be exactly one minute long.

And, in fact, they're even easier if there are no end times at all (a row is implicitly ended by the existence of a subsequent row).

This does mean that for a value to revert to "blank", an explicit row needs to be added forcing that blank.

For example...

Would become...

The first part of my code is to do just that, overly carefully...

• This can be completely skipped if there are no gaps in the data
• Your example data has no gaps, so this code (and the END_TS column) may be skippable

WITH
  fix_end AS
(
  SELECT
    MASTER_WORK_ORDR_ID,
    START_TS,
    DATEADD(microsecond, 1, NULLIF(END_TS,'9999-12-31 23:59:59.999999'))   AS END_TS,
    WORK_ORDR_ID_CTXT,
    WORK_ORDR_NUM
  FROM
    WORK_ORDR_IDENTIFIER_HIST
),
  unioned_terminations AS
(
  SELECT
    MASTER_WORK_ORDR_ID, START_TS, WORK_ORDR_ID_CTXT, WORK_ORDR_NUM
  FROM
    fix_end
  
  UNION ALL
  
  -- Generate rows where there are gaps in time for any `(MASTER_WORK_ORDR_ID, WORK_ORDR_ID_CTXT)`
  SELECT
    MASTER_WORK_ORDR_ID, END_TS, WORK_ORDR_ID_CTXT, ''
  FROM
    fix_end
  QUALIFY
    END_TS
    <
    LEAD(START_TS)
      OVER (
        PARTITION BY MASTER_WORK_ORDR_ID, WORK_ORDR_ID_CTXT
            ORDER BY START_TS
      )
)
SELECT * FROM unioned_terminations
  ORDER BY MASTER_WORK_ORDR_ID,WORK_ORDR_ID_CTXT,START_TS;

Once we have the data in that format (no overlaps, no gaps, rows end due to the existence of a following row) the data can then be pivotted, and LAST_VALUE() IGNORE NULLS used to fill the gaps.

SELECT
  MASTER_WORK_ORDR_ID,
  START_TS,
  LEAD(START_TS          )              OVER (PARTITION BY MASTER_WORK_ORDR_ID ORDER BY START_TS)   AS END_TS,
  LAST_VALUE(CALLID      ) IGNORE NULLS OVER (PARTITION BY MASTER_WORK_ORDR_ID ORDER BY START_TS)   AS CTXT_CALLID,
  LAST_VALUE(NPA         ) IGNORE NULLS OVER (PARTITION BY MASTER_WORK_ORDR_ID ORDER BY START_TS)   AS CTXT_NPA,
  LAST_VALUE(WFM         ) IGNORE NULLS OVER (PARTITION BY MASTER_WORK_ORDR_ID ORDER BY START_TS)   AS CTXT_WFM,
  LAST_VALUE(WFM_INTERNAL) IGNORE NULLS OVER (PARTITION BY MASTER_WORK_ORDR_ID ORDER BY START_TS)   AS CTXT_WFM_INTERNAL
FROM
(
  SELECT
    MASTER_WORK_ORDR_ID, START_TS, WORK_ORDR_ID_CTXT, WORK_ORDR_NUM
  FROM
    WORK_ORDR_IDENTIFIER_HIST
)
  AS src
/*
 * Or `unioned_terminations AS src` if there can be gaps in the data
 */
PIVOT   
(
  MAX(WORK_ORDR_NUM) 
    FOR WORK_ORDR_ID_CTXT IN
    (
       "NPA", "WFM_INTERNAL", "WFM", "CALLID" 
    ) 
)
    AS pvt
ORDER BY
  MASTER_WORK_ORDR_ID,
  START_TS
;

demo using SQL SERVER, as I don't know of a demo host for Teradata